# mulesoft Anypoint Platform Private Cloud Edition - Full Documentation --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Add a Product License Using Ops Center > For the full documentation index, see: https://docs.mulesoft.com/llms.txt When your product license expires, Ops Center displays a message stating that the system is under maintenance. It continues to display this message until you add a valid license. To add a new Anypoint Platform Private Cloud Edition license in Ops Center: 1. Sign into Ops Center. 2. Select the **Admin** tab. 3. Under **License**, add your new license in the text area. 4. Select **Update License**. ## See Also - [Ops Center](https://docs.mulesoft.com/private-cloud/3.2/managing-via-the-ops-center) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Adding Compatible Java Versions for Assets Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Add the compatible Java versions of assets for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) using the Maven Facade API (v1 and v2). The Java versions display in the Exchange UI, Exchange API, and GraphQL console for Anypoint Platform PCE environments. Only Java 8, 11, and 17 are supported. To add the Java version compatibility, in the POM file, add the properties section and set each Java version to either true or false. This example snippet shows the properties section: ```xml app connector, mulesoft false true true ``` In the example, Java 8 is set to `false`. Only version 11 and 17 appear for the asset version in the Exchange UI, Exchange API, and the GraphQL console. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Adding Compatible Java Versions for Assets Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Add the compatible Java versions of assets for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) using the Maven Facade API (v1 and v2). The Java versions display in the Exchange UI, Exchange API, and GraphQL console for Anypoint Platform PCE environments. Only Java 8, 11, and 17 are supported. To add the Java version compatibility, in the POM file, add the properties section and set each Java version to either true or false. This example snippet shows the properties section: ```xml app connector, mulesoft false true true ``` In the example, Java 8 is set to `false`. Only version 11 and 17 appear for the asset version in the Exchange UI, Exchange API, and the GraphQL console. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Adding Compatible Java Versions for Assets Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Add the compatible Java versions of assets for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) using the Maven Facade API (v1 and v2). The Java versions display in the Exchange UI, Exchange API, and GraphQL console for Anypoint Platform PCE environments. Only Java 8, 11, and 17 are supported. To add the Java version compatibility, in the POM file, add the properties section and set each Java version to either true or false. This example snippet shows the properties section: ```xml app connector, mulesoft false true true ``` In the example, Java 8 is set to `false`. Only version 11 and 17 appear for the asset version in the Exchange UI, Exchange API, and the GraphQL console. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Adding Compatible Java Versions for Assets Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Add the compatible Java versions of assets for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) using the Maven Facade API (v1 and v2). The Java versions display in the Exchange UI, Exchange API, and GraphQL console for Anypoint Platform PCE environments. Only Java 8, 11, and 17 are supported. To add the Java version compatibility, in the POM file, add the properties section and set each Java version to either true or false. This example snippet shows the properties section: ```xml app connector, mulesoft false true true ``` In the example, Java 8 is set to `false`. Only version 11 and 17 appear for the asset version in the Exchange UI, Exchange API, and the GraphQL console. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Allow Addresses for API Platform Proxies > For the full documentation index, see: https://docs.mulesoft.com/llms.txt By default, proxy requests are enabled for the domain name where the platform runs. > [!NOTE] For versions 3.2.2 and later, refer to Trusted Domains. To allow addresses in versions 3.2.0 and 3.2.1: 1. Log in to Ops Center and select **Kubernetes** on the left side-bar. 2. Select the `api-manager` namespace from the drop-down menu on the top, and then select **Configmaps**. Find the `api-platform-web-env` configmap and then click on `Edit Config Map`. 3. Select the **APW\_FEATURES\_PROXYWHITELIST** tab and ensure that value is set to `true`. (You’ll have to hover your mouse over the tabs to get a tooltip with the full name to find the right tab) If not, set it to `true` and then select `Apply`. 4. Select the **APW\_PROXYWHITELIST\_RULES** tab. 5. Write your rules separated by commas. The following example defines regular expressions that allow requests to be made to the `**.somewhere.com/**` and `**.somewhereelse.com/**` domains, where `*` is any part of a DNS name or URL: .\*\\.somewhere\\.com,.\*\\.somewhereelse\\.com 1. Select **Apply** to save changes to the `api-platform-web-env` config map. 2. Re-create the pods to ensure that each node in the cluster uses the most recent configuration: kubectl delete pod -n api-manager -l component=api-platform-web ## Trusted Domains Every HTTP request you make to an external resource goes through the API Console Proxy. For security reasons, you must register the domain as an allowed domain in Anypoint Platform. By default, proxy requests are blocked for external domains on authenticated requests. To add a new domain: 1. Log in to Anypoint Platform using an account that has the Organization Administrator permission. 2. In the navigation bar, click **Access Management**. 3. Click **Trusted domains**. 4. Click **Add domain**. The domain must not match an existing domain, must not be a domain or subdomain of MuleSoft or Anypoint Platform, and must be written in the format `exampledomain.com`. NOTE: Regular expressions are not supported. 5. Type the domain name and click **Add domain**. The domain is added to your list of trusted domains. You are redirected to the trusted domains page, where you can edit or delete your domains. If you register a domain on a parent organization, the domain is allowlisted for all of your business groups. ## Authentication When using the proxy, there are two authentication modes: - Strict Authentication: (Default) All requests that reach the proxy must be authenticated, which means that user token credentials must be included in the `ms-authorization` header of the request. > [!WARNING] If requests to external domains from the Exchange Public Portals go through the API > Console Proxy, they fail because no Anypoint user is involved in the call. - Non-Strict Authentication: API Console Proxy continues to be unauthenticated but it allows unauthenticated requests. The incoming request must provide either the user’s token or a reference to the Organization ID that the request belongs to. The Organization ID must be sent on the `ms-organization-id` header. To change the authentication method from strict to non-strict, follow these steps: 1. Log in to Ops Center and select **Kubernetes** on the left side-bar. 2. Select the `api-console-proxy` namespace from the drop-down menu on the top, and select **Configmaps**. 3. Locate the `service-env` configmap and then click on **Edit Config Map**. 4. Select the **STRICT\_AUTHENTICATION\_FF** tab and ensure that the value is set to `false`. (To find the right tab, hover your mouse over the tabs to get a tooltip with the full name.) 5. Click **Apply** to save changes to the `service-env` config map. 6. Recreate the pods to ensure that each node in the cluster uses the most recent configuration: kubectl delete pod -n api-console-proxy -l component=service --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Allowing Addresses for API Platform Proxies > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To allow addresses in Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0 and later, follow these steps: 1. From the api-manager/api-platform-web-env configmap verify if `APW_FEATURES_PROXYWHITELIST` is set to `true` by running the following command: `kubectl get cm -n api-manager api-platform-web-env -o jsonpath={.data.APW_FEATURES_PROXYWHITELIST} true` 2. If the response is not `true`, then set the value to `true` running the following command: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_FEATURES_PROXYWHITELIST": "true" }}` 3. Update the `APW_PROXYWHITELIST_RULES` in the same configmap by running the following command to patch the configmap with desired allow list rules: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_PROXYWHITELIST_RULES": ".**\\.somewhere\\.com,.**\\.somewhereelse\\.com" }}` The previous example shows how to allow requests to `.somewhere.com/` and `.somewhereelse.com/` domains, where `*` is any part of a DNS name or URL, then list the rules separated by commas. 4. Verify the patch was applied correctly. After confirmed, re-create the pods to ensure each node in the cluster uses the most recent configuration by running this command: `kubectl delete pod -n api-manager -l component=api-platform-web` ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Allowing Addresses for API Platform Proxies > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To allow addresses in Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0 and later, follow these steps: 1. From the api-manager/api-platform-web-env configmap verify if `APW_FEATURES_PROXYWHITELIST` is set to `true` by running the following command: `kubectl get cm -n api-manager api-platform-web-env -o jsonpath={.data.APW_FEATURES_PROXYWHITELIST} true` 2. If the response is not `true`, then set the value to `true` running the following command: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_FEATURES_PROXYWHITELIST": "true" }}` 3. Update the `APW_PROXYWHITELIST_RULES` in the same configmap by running the following command to patch the configmap with desired allow list rules: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_PROXYWHITELIST_RULES": ".**\\.somewhere\\.com,.**\\.somewhereelse\\.com" }}` The previous example shows how to allow requests to `.somewhere.com/` and `.somewhereelse.com/` domains, where `*` is any part of a DNS name or URL, then list the rules separated by commas. 4. Verify the patch was applied correctly. After confirmed, re-create the pods to ensure each node in the cluster uses the most recent configuration by running this command: `kubectl delete pod -n api-manager -l component=api-platform-web` ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Allowing Addresses for API Platform Proxies > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To allow addresses in Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0 and later, follow these steps: 1. From the api-manager/api-platform-web-env configmap verify if `APW_FEATURES_PROXYWHITELIST` is set to `true` by running the following command: `kubectl get cm -n api-manager api-platform-web-env -o jsonpath={.data.APW_FEATURES_PROXYWHITELIST} true` 2. If the response is not `true`, then set the value to `true` running the following command: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_FEATURES_PROXYWHITELIST": "true" }}` 3. Update the `APW_PROXYWHITELIST_RULES` in the same configmap by running the following command to patch the configmap with desired allow list rules: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_PROXYWHITELIST_RULES": ".**\\.somewhere\\.com,.**\\.somewhereelse\\.com" }}` The previous example shows how to allow requests to `.somewhere.com/` and `.somewhereelse.com/` domains, where `*` is any part of a DNS name or URL, then list the rules separated by commas. 4. Verify the patch was applied correctly. After confirmed, re-create the pods to ensure each node in the cluster uses the most recent configuration by running this command: `kubectl delete pod -n api-manager -l component=api-platform-web` ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Allowing Addresses for API Platform Proxies > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To allow addresses in Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0 and later, follow these steps: 1. From the api-manager/api-platform-web-env configmap verify if `APW_FEATURES_PROXYWHITELIST` is set to `true` by running the following command: `kubectl get cm -n api-manager api-platform-web-env -o jsonpath={.data.APW_FEATURES_PROXYWHITELIST} true` 2. If the response is not `true`, then set the value to `true` running the following command: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_FEATURES_PROXYWHITELIST": "true" }}` 3. Update the `APW_PROXYWHITELIST_RULES` in the same configmap by running the following command to patch the configmap with desired allow list rules: `kubectl patch cm -n api-manager api-platform-web-env -p '{"data":{"APW_PROXYWHITELIST_RULES": ".**\\.somewhere\\.com,.**\\.somewhereelse\\.com" }}` The previous example shows how to allow requests to `.somewhere.com/` and `.somewhereelse.com/` domains, where `*` is any part of a DNS name or URL, then list the rules separated by commas. 4. Verify the patch was applied correctly. After confirmed, re-create the pods to ensure each node in the cluster uses the most recent configuration by running this command: `kubectl delete pod -n api-manager -l component=api-platform-web` ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Analyze Business and API Data Using ELK > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Use the Elastic Stack (ELK stack) to analyze the business data and API analytics generated by Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). You can use Filebeat to process Anypoint Platform log files, insert them into an Elasticsearch database, and then analyze them with Kibana. The following example shows how API data appears in Kibana: ![kibana example](/mulesoft/current/en-us/private-cloud/3.2/_images/kibana-example.png) ## Prerequisites - Required software To export external analytics data to ELK, you must have the following software: - Mule runtime engine (Mule), version 3.8.4 or later - Anypoint Runtime Manager agent, version 1.7.0 or later - Elasticsearch, version 7.6.2 - Filebeat, version 7.6.2 - Kibana, version 7.6.2 - Ensure that you have installed and configured your Mule runtime engine instances. - Ensure that you have registered your Mule runtime engine instances with Runtime Manager. ## Edit the wrapper.conf Properties File In your Mule runtime installation, edit the following properties in the `conf/wrapper.conf` file: 1. Set the `analytics_enabled` property to `true`: wrapper.java.additional.=-Danypoint.platform.analytics\_enabled=true 2. Remove the URI value from the `analytics_base_uri` property: wrapper.java.additional.=-Danypoint.platform.analytics\_base\_uri= 3. Set the `on_prem` property to `true`: wrapper.java.additional.=-Danypoint.platform.on\_prem=true ## Enable Event Tracking To export business and event tracking data, you must enable event tracking in Runtime Manager: 1. From Anypoint Platform, select **Runtime Manager**. 2. Select **Servers**. 3. Select the row containing the server on which you want to configure ELK. 4. Select **Manager Server** and then select the **Plugins** tab. 5. In **Event Tracking**, select **Business Events** from the **Levels** drop-down menu. 6. Select the switch to enable event tracking on ELK. Runtime Manager displays the **ELK Integration** dialog. 7. Configure the log files as necessary. 8. Select **Apply**. ## Enable API Analytics To export API analytics, you must enable API analytics in Runtime Manager. 1. From Anypoint Platform, select **Runtime Manager**. 2. Select **Servers**. 3. Select the row containing the server on which you want to configure ELK. 4. Select the switch to enable API analytics and then select the switch next to **ELK**. 5. Configure the log files as necessary. 6. Select **Apply**. ## Install and Configure the Filebeat Agent On each server with Mule runtime instances, download and install Filebeat for your operating system. ## Configure the Mule Filebeat Module On each server with Mule runtime instances, download and expand the Filebeat Module archive from the following URL: https://s3.us-east-2.amazonaws.com/elk-integration/elk-integration-06-08-20/elk-integration.tar.gz 1. Add the absolute paths of each log file to the `manifest.yml` files: - Inside the `businessevents` folder, set the event log paths to the `var` variable of your operating system (OS): var: - name: paths default: - /var/mule/logs/events.log os.darwin: - /var/mule/logs/events.log os.windows: - /var/mule/logs/events.log - Inside the `apianalytics` folder, set the API analytics log paths to the `var` variable of your OS: var: - name: paths default: - /var/mule/logs/api-analytics-elk.log\* os.darwin: - /var/mule/logs/api-analytics-elk.log\* os.windows: - c:/programdata/mule/logs/api-analytics-elk.log\* 2. Copy the `mule` module folder to the `module` folder of your Filebeat installation. 3. Copy the `mule.yml` file to the `modules.d` folder of your Filebeat installation. 4. Start Elasticsearch and then Kibana. 5. Set up Filebeat and provision dashboards to Kibana. Inside your Filebeat installation folder, execute the following command. Replace the `${filebeat-installation-folder}` with the path to your Filebeat installation root folder, and change the Elasticsearch host if needed. *** /filebeat setup -e \\ -E output.elasticsearch.hosts="\['localhost:9200'\]" \\ -E output.elasticsearch.index="mule-%{+yyyy.MM.dd}" \\ -E setup.dashboards.enabled=true \\ -E setup.dashboards.directory=${filebeat-installation-folder}/module/mule/\_meta/kibana \\ -E setup.ilm.enabled=false \\ -E setup.template.name="mule" \\ -E setup.template.pattern="mule-\*" \\ -E setup.template.overwrite=true --- 6. Start Filebeat. Inside your Filebeat installation folder, execute the following command. Replace the `${filebeat-installation-folder}` with the path to your Filebeat installation root folder, and change the Elasticsearch host if needed. ./filebeat -e \\ -E output.elasticsearch.hosts="\['localhost:9200'\]" \\ -E output.elasticsearch.index="mule-%{+yyyy.MM.dd}" \\ -E setup.dashboards.enabled=true \\ -E setup.dashboards.directory=${filebeat-installation-folder}/module/mule/\_meta/kibana \\ -E setup.ilm.enabled=false \\ -E setup.template.name="mule" \\ -E setup.template.pattern="mule-\*" \\ -E setup.template.overwrite=true ## Configure Kibana to Load an Index After installing Filebeat and Elasticsearch, you must configure Kibana to consume data from Anypoint Platform. MuleSoft provides a default Kibana configuration that you can use to analyze business and API data. This configuration includes dashboards, searches, and visualizations. 1. Configure an index pattern You must create an Elasticsearch index for Anypoint Platform data: 1. Generate an initial set of data. This is required for Kibana to be able to recognize the index to be created. To send a request to a test API to generate an initial set of data: 1. In the Kibana management console, create an index pattern with `mule-*` as the value. 2. Select **@timestamp** in **Time Filter field name**. 3. In **Show advanced options**, set **mule-index** as the value in the **Custom index pattern ID** field. 2. Create the index. You can now access all the Mule dashboards in the **Dashboard** tab. --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Anypoint Monitoring on Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) includes support for Anypoint Monitoring. Anypoint Monitoring is a suite of tools that provides feedback from Mule flows and components in your application network. Anypoint Monitoring is an add-on to Anypoint Platform PCE. To use Anypoint Monitoring with Anypoint Platform PCE, contact MuleSoft Professional Services. Using Anypoint Monitoring with Anypoint Platform PCE requires the "Monitoring for On-Premises Management" subscription. This subscription entitles you to use two monitoring instances. A monitoring instance is an implementation of the 3-node Anypoint Monitoring configuration described in [Supported Configurations for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/supported-cluster-config). Each monitoring instance is dedicated to a single Anypoint Platform PCE installation. To use Anypoint Monitoring with more than two Anypoint Platform PCE installations, you must purchase additional monitoring subscriptions. Contact MuleSoft Professional Services for more information. ## Supported Anypoint Monitoring Features
FeatureGold and Platinum Subscription FeaturesTitanium Subscription FeaturesAnypoint Platform PCE Features

Dedicated infrastructure for data processing and storage

No

Yes

Yes

Custom metrics

No

Yes

Yes

Built-in Application Dashboards

More than 80 out-of-the-box charts covering metrics in the following categories: Overview, Inbound, Outbound, Performance, Failures, JVM, and Infrastructure

Yes

Yes

Yes

Advanced metrics for Inbound and Outbound charts

No

Yes

Yes

Flows charts

No

Yes

Yes

Connectors charts

No

Yes

Yes

Anypoint Visualizer application network integration

No

Yes

Yes

Data retention for more than 30 days

No

Yes

Yes

Built-in API Dashboards

16 out-of-the-box charts covering metrics in the following categories: Overview, Requests, Failures, Performance, and Client Application

Yes

Yes

Yes

Data retention for more than 30 days

No

Yes

Yes

Custom Dashboards

Basic mode

Yes

Yes

Yes

Advanced mode

Yes, with limited functionality

Yes

Yes

Advanced mode with additional features

No

Yes

Yes

Alerts

Basic alerts

Yes

Yes, with higher limits

Yes (Applications)

Advanced alerts

No

Yes

Yes (Applications and API)

Log Management

Log search

No

Yes

No

Log points

No

Yes

No

Raw data

No

Yes

No

Additional Anypoint Monitoring features

Anypoint Monitoring Archive API

No

Yes

No

API Functional Monitoring

No

No

No

Reports

No

Yes

Yes

Enhanced Support

Enhanced Support

No

Yes

No

### Unsupported Anypoint Monitoring Features Anypoint Platform PCE does not the support the following Anypoint Monitoring features: - Support for Mule 3.8.x - Log management As an alternative, use your server’s logging capabilties or a third-party vendor like Splunk. - Increased storage capacity This capacity is not required because the Archive API is not supported. - Archive API This API is not applicable because log management is not supported. - API Functional Monitoring As an alternative, use a third-party vendor, such as Pingdom. - The **Tools** option in the left-hand menu is not available. ## Requirements Anypoint Monitoring has disk and network port requirements in addition to the Anypoint Platform PCE requirements. For information, see [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware). ## Backup and Restore Anypoint Monitoring backup and restore are outside the normal gravity backup and restore mechanism, and require the additional procedures. For information, see [Configure Backup and Restore for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/backup-and-disaster-recovery). ## Uninstall Anypoint Monitoring If you want to uninstall the Anypoint Monitoring add-on, contact MuleSoft services. ## See Also - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Configure Backup and Restore for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/backup-and-disaster-recovery) - [Anypoint Monitoring Overview](https://docs.mulesoft.com/monitoring/) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Anypoint Platform PCE Overview > For the full documentation index, see: https://docs.mulesoft.com/llms.txt With Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), you run and manage Mule applications on your local servers using your required security policies. With Anypoint Platform PCE, the control plane functionality is locally available and does not require interaction with external systems or public internet connectivity. You configure and run Anypoint Platform software and maintain all data storage, processing, and transmission. Refer to the following table for components included in Anypoint Platform PCE. Anypoint Platform PCE uses Docker and Kubernetes to provide built-in high availability and scalability. You can install the platform without understanding Docker or Kubernetes, but you must understand how they work to manage and troubleshoot the platform. For information about a previous version of Anypoint Platform PCE, select the version in the documentation sidebar. ## Anypoint Platform PCE Components The following table provides component information for Anypoint Platform PCE:
ComponentIncludedNot IncludedDescription

Anypoint Code Builder

X

Enables you to create API specifications, implementations, and integration projects.

Anypoint Runtime Manager

X

Enables you to track and manage your deployed Mule applications.

Runtime Fabric Management (including Security Edge and Tokenization)

X

Automates the deployment and orchestration of customer-hosted Mule applications and API gateways.

Anypoint API Manager

X

Enables you to manage your registered APIs.

Secrets Manager

X

Stores and manages secrets for supported Anypoint Platform services.

Anypoint Exchange

X

Enables your organization to share various assets for reuse across integration projects.

Access Management

X

Enables you to create your Anypoint Platform account or configure external federated identity.

Anypoint Analytics - Insights

X

Provides in-depth visibility into business transactions and events on your Mule apps deployed to CloudHub and on-premises servers.

Anypoint Analytics - Anypoint Visualizer

X

Enables you to explore your application network via an application network graph.

Anypoint Analytics - API

X

Provides insight into how your APIs are being used and how they are performing.

Anypoint Monitoring

X

Provides visibility into integrations, Mule flows, and components in your application network.

Object Store

X

Object Store v2 allows CloudHub applications to store data and states across batch processes, Mule components, and applications, either from within an application or by using the Object Store REST API.

Anypoint MQ

X

A multi-tenant cloud messaging service that enables you to perform advanced asynchronous messaging scenarios between tenant applications.

API Testing

X

Provides a testing framework that enables you to easily build automated tests for your integrations and APIs.

For Anypoint Platform PCE version 3.2, the following policies are included with API Manager:
PolicyVersion

Basic authentication: Simple

1.2.2

Basic Authentication - LDAP

1.3.1

Client ID Enforcement

1.2.4

Cross-Origin Resource Sharing

1.2.1

Detokenization

1.1.2

Tokenization

1.1.2

Header Injection

1.2.1

Header Removal

1.0.3

HTTP Caching

1.0.5

IP Blocklist

1.0.0

IP Allowlist

1.0.0

JSON Threat Protection

1.1.5

JWT Validation

1.3.0

Message Logging

1.0.0

Mule OAuth 2.0 Access Token

1.5.1

OpenAM OAuth 2.0 Token Enforcement

1.5.1

OpenID Connect OAuth 2.0 Token Enforcement

1.5.1

PingFederate Access Token Enforcement

1.5.1

Rate Limiting

1.3.6

Rate Limiting - SLA based

1.2.8

Spike Control

1.1.5

XML Threat Protection

1.1.4

For additional information about policies, refer to [Policies](https://docs.mulesoft.com/gateway/latest/gateway-secure-apis). ## Mule Runtime Version Compatibility - Anypoint Platform PCE is compatible with Mule runtime engine v3.9 and later. - Anypoint Platform PCE supports the Mule runtime engine LTS and Edge release channels. The recommended customer profile is LTS. For more information, visit [Edge and LTS Releases for Mule](https://docs.mulesoft.com/release-notes/mule-runtime/lts-edge-release-cadence). ## Features Shipped with Anypoint Platform PCE Version 3.2 Refer to [Anypoint Platform Private Cloud Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-release-notes) for detailed release information. Anypoint Platform PCE version 3.2 contains the following features:
ComponentFeature DescriptionFeature Version

Exchange

API Metadata Service

v0.7.636

Asset Manager Service

v1.0.504

Asset Portal Service

v0.6.1192

Assets Reviews Service

v0.8.71

Custom Assets Facade Service

v0.9.246

Customization Service

v0.6.214

Dependency Resolver Service

v0.6.469

File Service

v1.1.47

File Upload Service

v0.6.86

Graph Node Service

v0.2.172

Graph Service

v0.8.584

Maven Facade Service

v0.7.757

Search Service

v0.8.250

Exchange UI

v0.8.4803

XAPI Service

v0.8.2127

Connectivity

REST Connect Service

v1.12.40

API Platform

API Platform API

v2.7.792

API Platform Web

v1.15.117

API Manager Web

v1.5.1031

Contract Caching Service

v1.1.158

API Proxy Builder Service

v1.2.123

RAML Parser Service

v0.3.102

Proxies XAPI Service

v1.2.206

API Designer

API Designer XAPI

v3.41.0

API Designer UI

v2.44.0

Mocking Service

Mocking Service

v2.41.0

Mozart

Design Center UI

v2.0.143

Mozart XAPI

v2.0.232

Mozart Project Manager Service

v3.0.45

Runtime Manager

Cloudhub MCM

v2.17.155

Cloudhub UI

1.100

AMC UI REST Facade

v2.12.75

Hybrid REST

v2.31.269

Metadata Manager

1.9.49

Platform Alerts

1.10.59

Mule Message Processor

v2.0.37

Platform Messages

1.7.58

Access Management

Authentication Server

v1.2.48

Core Services UI

v0.81.x

Anypoint Sign In

v0.81.x

Audit Log

v0.81.x

Landing Page UI

v0.81.x

Navbar

v0.81.x

Anypoint Platform PCE

Anypoint Platform PCE Packaging

v3.2.0

API Modeling Framework (AMF)

AMF Service

v1.2.49

VCS

VCS

v0.24+

VCS Git Proxy

v0.2.132+

Anypoint Studio

Studio

v7.15.0

## Add-On Features Supported by Anypoint Platform PCE Version 3.2
FeatureVersion

Anypoint Monitoring

v0.6.2.0

## See Also - [Supported Configurations](https://docs.mulesoft.com/private-cloud/3.2/supported-cluster-config) - [Product Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/install-checklist) - [Runtime Manager](https://docs.mulesoft.com/runtime-manager/) - [Anypoint Exchange](https://docs.mulesoft.com/exchange/) - [API Manager](https://docs.mulesoft.com/api-manager/latest/) - [Docker](https://www.docker.com/) - [Kubernetes](https://kubernetes.io/) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Anypoint Platform PCE Overview > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Run and manage Mule apps on your local servers following your security policies by using Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). Anypoint Platform PCE offers local MuleSoft control plane functionality, with no external system interactions or internet connectivity required, maintaining all data storage, processing, and transmission. Anypoint Platform PCE uses Kubernetes for built-in high availability and scalability. While you can install it without deep Kubernetes knowledge, understanding these tools is essential for management and troubleshooting. Anypoint Platform PCE version 4.0 runs on enterprise-ready Kubernetes Red Hat OpenShift, Amazon EKS Anywhere (EKS-A), and Rancher Kubernetes Engine (RKE2). Additionally, it supports Java 17, in addition to Java 8 and 11. For information about a previous version of Anypoint Platform PCE, select the version in the documentation sidebar. ## Before You Begin ### Upgrade to Anypoint Platform PCE version 4.0 If you are upgrading from version 3.2.3 of Anypoint Platform PCE to version 4.0, see [Upgrading Anypoint Platform PCE 3.2.3 to 4.0](https://docs.mulesoft.com/private-cloud/4.0/upgrade). > [!NOTE] If you’re using a version of Anypoint Platform PCE earlier than 3.2.3, upgrade to version > 3.2.3 first. Then proceed with upgrading to version 4.0. ### Install Anypoint Platform PCE 4.0 If you are installing Anypoint Platform PCE version 4.0 for the first time, see [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/install-workflow). To install Anypoint Platform PCE on OpenShift, see [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/4.0/install-openshift). ## Anypoint Platform PCE Components The following table provides component information for Anypoint Platform PCE:
ComponentIncludedDescription

Access Management

Yes

Enables you to create your Anypoint Platform account or configure external federated identity.

Anypoint API Manager

Yes

Enables you to manage your registered APIs.

Anypoint Analytics - API

No

Provides insight into how your APIs are being used and how they are performing.

Anypoint Analytics - Anypoint Visualizer

No

Enables you to explore your application network via an application network graph.

Anypoint Analytics - Insights

No

Provides in-depth visibility into business transactions and events on your Mule apps deployed to CloudHub and on-premises servers.

Anypoint Code Builder

No

Enables you to create API specifications, implementations, and integration projects.

Anypoint Exchange

Yes

Enables your organization to share various assets for reuse across integration projects.

Anypoint Monitoring

No

Provides visibility into integrations, Mule flows, and components in your application network.

Anypoint MQ

No

A multi-tenant cloud messaging service that enables you to perform advanced asynchronous messaging scenarios between tenant applications.

Anypoint Runtime Fabric Management

No

Automates the deployment and orchestration of customer-hosted Mule applications and API gateways.

Anypoint Runtime Manager

Yes

Enables you to track and manage your deployed Mule applications.

API Testing

No

Provides a testing framework that enables you to easily build automated tests for your integrations and APIs.

Object Store

No

Object Store v2 allows CloudHub applications to store data and states across batch processes, Mule components, and applications, either from within an application or by using the Object Store REST API.

Secrets Manager

No

Stores and manages secrets for supported Anypoint Platform services.

## Mule Runtime Version Compatibility - Anypoint Platform PCE is compatible with Mule runtime engine versions 4.3 - 4.6. Use version 4.6 LTS to get the best performance from all Anypoint Platform connectors. - Anypoint Platform PCE supports the Mule runtime engine LTS and Edge release channels. The recommended customer profile is LTS. For more information, visit [Edge and LTS Releases for Mule](https://docs.mulesoft.com/release-notes/mule-runtime/lts-edge-release-cadence). ## Java Version Compatibility Anypoint Platform PCE is compatible with Java 17, 11, and 8. ## Features Shipped with Anypoint Platform PCE Version 4.0 Refer to [Anypoint Platform Private Cloud Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-release-notes) for detailed release information. Anypoint Platform PCE version 4.0 contains the following features:
ComponentFeature DescriptionFeature Version

Access Management

Audit Log

v5.1.45

Authentication Server

v1.2.460

Authentication Server UI

v2.1.240-56.develop

Anypoint Sign In

v3.2.143

Anypoint Roles Client

v8.1.1573

ARC Informer

v1.0.374

API Designer

API Designer DC Projects Service

v3.13.1

API Designer DC XAPI Service

v2.11.2

API Designer DC UI

v2.10.5

API Manager

API Manager API DB Migrations

v1.0.451

API Manager API Platform Web

v1.15.204

API Manager API Proxy Builder Service

v1.3.69

API Manager API

v2.31.5

API Manager Contract Caching DB Migrations

v1.0.238

API Manager Contract Caching Service

v1.1.418

API Manager Proxies XAPI Service

v1.24.0

API Manager RAML Parser Service

v0.3.159

API Manager Web

v1.29.1

API Modeling Framework (AMF)

AMF Service

v1.4.61

Connectivity

REST Connect Service

v1.14.53

Exchange

Asset Manager DB Migrations

v2.0.20

Asset Manager Service

v1.0.808

Asset Portals DB Migrations

v2.0.25

Asset Portals Service

v0.6.1503

Content Provider

v1.0.12

Custom Assets Facade Service

v0.9.436

Customization Service

v0.6.214

Dependency Resolver Service

v0.8.27

File Service

v1.1.47

File Upload Service

v1.1.226

Maven Facade Service

v0.8.47

Region Synchronizer Service

v0.2.147

Exchange API Metadata DB Migrations

v1.7.324

Exchange API Metadata Service

v0.7.954

Exchange Asset Reviews DB Migrations

v1.6.359

Exchange Asset Reviews Service

v0.8.353

Exchange Customization DB Migrations

v1.0.264

Exchange Customization Service

v0.6.576

Exchange Customization Service

v0.6.576

Exchange Graph Node Service

v0.2.563

Exchange Search DB Migrations

v1.7.283

Exchange Search Service

v1.0.90

Exchange UI

v1.0.126

Exchange XAPI Service

v0.8.2717

Mocking Service

Mocking Service

v2.68.6

Runtime Manager

Application Manager DB Migration

v1.0.92

Application Manager Service

v2.33.0

Atlas Service

v1.9.109

Mule Agent

v2.6.6

MCM Service

v2.17.322

Runtime Services

Anypoint Flash Platform Messages DB Migrations

v1.1.31

Platform Messages

v2.2.21

Mule Message Processor

v3.2.17

## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) - [Upgrading Anypoint Platform PCE v3.2.3 to v4.0.0](https://docs.mulesoft.com/private-cloud/4.0/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Anypoint Platform PCE Overview > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Run and manage Mule apps on your local servers following your security policies by using Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). Anypoint Platform PCE offers local MuleSoft control plane functionality, with no external system interactions or internet connectivity required, maintaining all data storage, processing, and transmission. Anypoint Platform PCE uses Kubernetes for built-in high availability and scalability. While you can install it without deep Kubernetes knowledge, understanding these tools is essential for management and troubleshooting. Anypoint Platform PCE version 4.x runs on enterprise-ready Kubernetes Red Hat OpenShift, Amazon EKS Anywhere (EKS-A), and Rancher Kubernetes Engine (RKE2). Additionally, it supports Java 17, in addition to Java 8 and 11. For information about a previous version of Anypoint Platform PCE, select the version in the documentation sidebar. ## Before You Begin ### Upgrade to Anypoint Platform PCE version 4.X If you are upgrading from version 3.2.3 of Anypoint Platform PCE to version 4.X, see [Upgrading Anypoint Platform PCE 3.2.3 to 4.x](https://docs.mulesoft.com/private-cloud/4.1/upgrade). > [!NOTE] If you’re using a version of Anypoint Platform PCE earlier than 3.2.3, upgrade to version > 3.2.3 first. Then proceed with upgrading to version 4.x. ### Install Anypoint Platform PCE 4.X If you are installing Anypoint Platform PCE version 4.x for the first time, see [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/install-workflow). To install Anypoint Platform PCE on OpenShift, see [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/4.1/install-openshift). ## Anypoint Platform PCE Components This table provides component information for Anypoint Platform PCE:
ComponentIncludedDescription

Access Management

Yes

Enables you to create your Anypoint Platform account or configure external federated identity.

Anypoint API Manager

Yes

Enables you to manage your registered APIs.

Anypoint Analytics - API

No

Provides insight into how your APIs are being used and how they are performing.

Anypoint Analytics - Anypoint Visualizer

Yes

Enables you to explore your application network via an application network graph.

Anypoint Analytics - Insights

No

Provides in-depth visibility into business transactions and events on your Mule apps deployed to CloudHub and on-premises servers.

Anypoint Code Builder

No

Enables you to create API specifications, implementations, and integration projects.

Anypoint Exchange

Yes

Enables your organization to share various assets for reuse across integration projects.

Anypoint Monitoring

No

Provides visibility into integrations, Mule flows, and components in your application network.

Anypoint MQ

No

A multi-tenant cloud messaging service that enables you to perform advanced asynchronous messaging scenarios between tenant applications.

Anypoint Runtime Fabric Management

No

Automates the deployment and orchestration of customer-hosted Mule applications and API gateways.

Anypoint Runtime Manager

Yes

Enables you to track and manage your deployed Mule applications.

API Testing

No

Provides a testing framework that enables you to easily build automated tests for your integrations and APIs.

Object Store

No

Object Store v2 allows CloudHub applications to store data and states across batch processes, Mule components, and applications, either from within an application or by using the Object Store REST API.

Secrets Manager

No

Stores and manages secrets for supported Anypoint Platform services.

## Mule Runtime Version Compatibility - Anypoint Platform PCE is compatible with Mule runtime engine versions 4.3 - 4.6. Use version 4.6 LTS to get the best performance from all Anypoint Platform connectors. - Anypoint Platform PCE supports the Mule runtime engine LTS and Edge release channels. The recommended customer profile is LTS. For more information, visit [Edge and LTS Releases for Mule](https://docs.mulesoft.com/release-notes/mule-runtime/lts-edge-release-cadence). ## Java Version Compatibility Anypoint Platform PCE is compatible with Java 17, 11, and 8. ## Features Shipped with Anypoint Platform PCE Version 4.X Refer to [Anypoint Platform Private Cloud Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-release-notes) for detailed release information. Anypoint Platform PCE version 4.1 contains these features:
ComponentFeature DescriptionFeature Version

Access Management

Audit Log

v5.1.67

Authentication Server

v1.2.548

Authentication Server UI

v2.1.269-8.develop

Anypoint Sign In

v3.2.161

Anypoint Roles Client

v8.1.1573

ARC Informer

1.0.374

API Designer

API Designer DC Projects Service

v3.18.10

API Designer DC XAPI Service

v2.15.7

API Designer DC UI

v2.75.8

API Manager

API Manager API DB Migrations

v1.0.500

API Manager API Platform Web

v1.15.226

API Manager API Proxy Builder Service

v1.3.92

API Manager API

v2.31.26

API Manager Contract Caching DB Migrations

v1.0.259

API Manager Contract Caching Service

v1.1.459

API Manager Proxies XAPI Service

v1.24.11

API Manager RAML Parser Service

v0.3.173

API Manager Web

v1.29.13

API Modeling Framework (AMF)

AMF Service

v1.4.66

Connectivity

REST Connect Service

v1.14.57

Exchange

Asset Manager DB Migrations

v2.0.51

Asset Manager Service

v1.0.885

Asset Portals DB Migrations

v2.0.57

Asset Portals Service

v0.6.1576

Content Provider

v1.0.29

Custom Assets Facade Service

v1.0.7

Customization Service

v0.6.658

Dependency Resolver Service

v0.8.77

File Service

v1.1.288

File Upload Service

v1.1.226

Maven Facade Service

v0.8.109

Region Synchronizer Service

v0.2.163

Exchange API Metadata DB Migrations

v2.0.40

Exchange API Metadata Service

v1.1.1

Exchange Asset Reviews DB Migrations

v2.0.36

Exchange Asset Reviews Service

v0.8.435

Exchange Customization DB Migrations

2.0.31

Exchange Customization Service

0.6.658

Exchange Graph Node Service

0.2.664

Exchange Search DB Migrations

2.0.38

Exchange Search Service

v1.0.215

Exchange UI

v1.0.265

Exchange XAPI Service

v0.8.2848

Mocking Service

v2.77.0-17.pce

Runtime Manager

Application Manager DB Migration

v1.2.11

Application Manager Service

v2.33.51

Atlas Service

v1.9.121

Mule Agent

v2.6.6

Jafar Service

0.3.167

MCM Service

v2.17.347

Runtime Services

Anypoint Flash Platform Messages DB Migrations

1.1.43

Platform Messages

v2.2.43

Mule Message Processor

v3.2.40

## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) - [Upgrading Anypoint Platform PCE v3.2.5 to v4.1.x](https://docs.mulesoft.com/private-cloud/4.1/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Anypoint Platform PCE Overview > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Run and manage Mule apps on your local servers following your security policies by using Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). Anypoint Platform PCE offers local MuleSoft control plane functionality, with no external system interactions or internet connectivity required, maintaining all data storage, processing, and transmission. Anypoint Platform PCE uses Kubernetes for built-in high availability and scalability. While you can install it without deep Kubernetes knowledge, understanding these tools is essential for management and troubleshooting. Anypoint Platform PCE version 4.x runs on enterprise-ready Kubernetes Red Hat OpenShift and Rancher Kubernetes Engine (RKE2). Additionally, it supports Java 17, in addition to Java 8 and 11. For information about a previous version of Anypoint Platform PCE, select the version in the documentation sidebar. ## Before You Begin ### Upgrade to Anypoint Platform PCE version 4.X If you are upgrading from version 3.2.3 of Anypoint Platform PCE to version 4.X, see [Upgrading Anypoint Platform PCE 3.2.3 to 4.x](https://docs.mulesoft.com/private-cloud/4.2/upgrade). > [!NOTE] If you’re using a version of Anypoint Platform PCE earlier than 3.2.3, upgrade to version > 3.2.3 first. Then proceed with upgrading to version 4.x. ### Install Anypoint Platform PCE 4.X If you are installing Anypoint Platform PCE version 4.x for the first time, see [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/install-workflow). To install Anypoint Platform PCE on OpenShift, see [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/4.2/install-openshift). ## Anypoint Platform PCE Components This table provides component information for Anypoint Platform PCE:
ComponentIncludedDescription

Access Management

Yes

Enables you to create your Anypoint Platform account or configure external federated identity.

Anypoint API Manager

Yes

Enables you to manage your registered APIs.

Anypoint Analytics - API

No

Provides insight into how your APIs are being used and how they are performing.

Anypoint Analytics - Anypoint Visualizer

Yes

Enables you to explore your application network via an application network graph.

Anypoint Analytics - Insights

No

Provides in-depth visibility into business transactions and events on your Mule apps deployed to CloudHub and on-premises servers.

Anypoint Code Builder

Yes

Enables you to create API specifications, implementations, and integration projects.

Anypoint Exchange

Yes

Enables your organization to share various assets for reuse across integration projects.

Anypoint Monitoring

No

Provides visibility into integrations, Mule flows, and components in your application network.

Anypoint MQ

No

A multi-tenant cloud messaging service that enables you to perform advanced asynchronous messaging scenarios between tenant applications.

Anypoint Runtime Fabric Management

No

Automates the deployment and orchestration of customer-hosted Mule applications and API gateways.

Anypoint Runtime Manager

Yes

Enables you to track and manage your deployed Mule applications.

API Testing

No

Provides a testing framework that enables you to easily build automated tests for your integrations and APIs.

Object Store

No

Object Store v2 allows CloudHub applications to store data and states across batch processes, Mule components, and applications, either from within an application or by using the Object Store REST API.

Secrets Manager

No

Stores and manages secrets for supported Anypoint Platform services.

## Mule Runtime Version Compatibility - Anypoint Platform PCE is compatible with Mule runtime engine versions 4.3 - 4.6. Use version 4.6 LTS to get the best performance from all Anypoint Platform connectors. - Anypoint Platform PCE supports the Mule runtime engine LTS and Edge release channels. The recommended customer profile is LTS. For more information, visit [Edge and LTS Releases for Mule](https://docs.mulesoft.com/release-notes/mule-runtime/lts-edge-release-cadence). ## Java Version Compatibility Anypoint Platform PCE is compatible with Java 17, 11, and 8. ## Features Shipped with Anypoint Platform PCE Version 4.X Refer to [Anypoint Platform Private Cloud Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-release-notes) for detailed release information. Anypoint Platform PCE version 4.2 contains these features:
ComponentFeature DescriptionFeature Version

Access Management

Audit Log

v5.1.67

Authentication Server

v1.2.548

Authentication Server UI

v2.1.269-8.develop

Anypoint Sign In

v3.2.161

Anypoint Roles Client

v8.1.1573

ARC Informer

1.0.374

API Designer

API Designer DC Projects Service

v3.18.10

API Designer DC XAPI Service

v2.15.7

API Designer DC UI

v2.75.8

API Manager

API Manager API DB Migrations

v1.0.500

API Manager API Platform Web

v1.15.226

API Manager API Proxy Builder Service

v1.3.92

API Manager API

v2.31.26

API Manager Contract Caching DB Migrations

v1.0.259

API Manager Contract Caching Service

v1.1.459

API Manager Proxies XAPI Service

v1.24.11

API Manager RAML Parser Service

v0.3.173

API Manager Web

v1.29.13

API Modeling Framework (AMF)

AMF Service

v1.4.66

Connectivity

REST Connect Service

v1.14.57

Exchange

Asset Manager DB Migrations

v2.0.51

Asset Manager Service

v1.0.885

Asset Portals DB Migrations

v2.0.57

Asset Portals Service

v0.6.1576

Content Provider

v1.0.29

Custom Assets Facade Service

v1.0.7

Customization Service

v0.6.658

Dependency Resolver Service

v0.8.77

File Service

v1.1.288

File Upload Service

v1.1.226

Maven Facade Service

v0.8.109

Region Synchronizer Service

v0.2.163

Exchange API Metadata DB Migrations

v2.0.40

Exchange API Metadata Service

v1.1.1

Exchange Asset Reviews DB Migrations

v2.0.36

Exchange Asset Reviews Service

v0.8.435

Exchange Customization DB Migrations

2.0.31

Exchange Customization Service

0.6.658

Exchange Graph Node Service

0.2.664

Exchange Search DB Migrations

2.0.38

Exchange Search Service

v1.0.215

Exchange UI

v1.0.265

Exchange XAPI Service

v0.8.2848

Mocking Service

v2.77.0-17.pce

Runtime Manager

Application Manager DB Migration

v1.2.11

Application Manager Service

v2.33.51

Atlas Service

v1.9.121

Mule Agent

v2.6.6

Jafar Service

0.3.167

MCM Service

v2.17.347

Runtime Services

Anypoint Flash Platform Messages DB Migrations

1.1.43

Platform Messages

v2.2.43

Mule Message Processor

v3.2.40

## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.2/install-checklist) - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) - [Migrating Anypoint Platform PCE v3.X to v4.X](https://docs.mulesoft.com/private-cloud/4.2/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Anypoint Platform PCE Overview > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Run and manage Mule apps on your local servers following your security policies by using Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). Anypoint Platform PCE offers local MuleSoft control plane functionality, with no external system interactions or internet connectivity required, maintaining all data storage, processing, and transmission. Anypoint Platform PCE uses Kubernetes for built-in high availability and scalability. While you can install it without deep Kubernetes knowledge, understanding these tools is essential for management and troubleshooting. Anypoint Platform PCE version 4.x runs on enterprise-ready Kubernetes Red Hat OpenShift and Rancher Kubernetes Engine (RKE2). Additionally, it supports Java 17, in addition to Java 8 and 11. For information about a previous version of Anypoint Platform PCE, select the version in the documentation sidebar. ## Before You Begin ### Upgrade to Anypoint Platform PCE version 4.X If you are upgrading from version 3.2.3 of Anypoint Platform PCE to version 4.X, see [Upgrading Anypoint Platform PCE 3.2.3 to 4.x](https://docs.mulesoft.com/private-cloud/latest/upgrade). > [!NOTE] If you’re using a version of Anypoint Platform PCE earlier than 3.2.3, upgrade to version > 3.2.3 first. Then proceed with upgrading to version 4.x. ### Install Anypoint Platform PCE 4.X If you are installing Anypoint Platform PCE version 4.x for the first time, see [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/install-workflow). To install Anypoint Platform PCE on OpenShift, see [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/latest/install-openshift). ## Anypoint Platform PCE Components This table provides component information for Anypoint Platform PCE:
ComponentIncludedDescription

Access Management

Yes

Enables you to create your Anypoint Platform account or configure external federated identity.

Anypoint API Manager

Yes

Enables you to manage your registered APIs.

Anypoint Analytics - API

No

Provides insight into how your APIs are being used and how they are performing.

Anypoint Analytics - Anypoint Visualizer

Yes

Enables you to explore your application network via an application network graph.

Anypoint Analytics - Insights

No

Provides in-depth visibility into business transactions and events on your Mule apps deployed to CloudHub and on-premises servers.

Anypoint Code Builder

Yes

Enables you to create API specifications, implementations, and integration projects.

Anypoint Exchange

Yes

Enables your organization to share various assets for reuse across integration projects.

Anypoint Monitoring

No

Provides visibility into integrations, Mule flows, and components in your application network.

Anypoint MQ

No

A multi-tenant cloud messaging service that enables you to perform advanced asynchronous messaging scenarios between tenant applications.

Anypoint Runtime Fabric Management

No

Automates the deployment and orchestration of customer-hosted Mule applications and API gateways.

Anypoint Runtime Manager

Yes

Enables you to track and manage your deployed Mule applications.

API Testing

No

Provides a testing framework that enables you to easily build automated tests for your integrations and APIs.

Object Store

No

Object Store v2 allows CloudHub applications to store data and states across batch processes, Mule components, and applications, either from within an application or by using the Object Store REST API.

Secrets Manager

No

Stores and manages secrets for supported Anypoint Platform services.

## Mule Runtime Version Compatibility - Anypoint Platform PCE is compatible with Mule runtime engine versions 4.3 - 4.6. Use version 4.6 LTS to get the best performance from all Anypoint Platform connectors. - Anypoint Platform PCE supports the Mule runtime engine LTS and Edge release channels. The recommended customer profile is LTS. For more information, visit [Edge and LTS Releases for Mule](https://docs.mulesoft.com/release-notes/mule-runtime/lts-edge-release-cadence). ## Java Version Compatibility Anypoint Platform PCE is compatible with Java 17, 11, and 8. ## Features Shipped with Anypoint Platform PCE Version 4.X Refer to [Anypoint Platform Private Cloud Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-release-notes) for detailed release information. Anypoint Platform PCE version 4.2 contains these features:
ComponentFeature DescriptionFeature Version

Access Management

Audit Log

v5.1.67

Authentication Server

v1.2.548

Authentication Server UI

v2.1.269-8.develop

Anypoint Sign In

v3.2.161

Anypoint Roles Client

v8.1.1573

ARC Informer

1.0.374

API Designer

API Designer DC Projects Service

v3.18.10

API Designer DC XAPI Service

v2.15.7

API Designer DC UI

v2.75.8

API Manager

API Manager API DB Migrations

v1.0.500

API Manager API Platform Web

v1.15.226

API Manager API Proxy Builder Service

v1.3.92

API Manager API

v2.31.26

API Manager Contract Caching DB Migrations

v1.0.259

API Manager Contract Caching Service

v1.1.459

API Manager Proxies XAPI Service

v1.24.11

API Manager RAML Parser Service

v0.3.173

API Manager Web

v1.29.13

API Modeling Framework (AMF)

AMF Service

v1.4.66

Connectivity

REST Connect Service

v1.14.57

Exchange

Asset Manager DB Migrations

v2.0.51

Asset Manager Service

v1.0.885

Asset Portals DB Migrations

v2.0.57

Asset Portals Service

v0.6.1576

Content Provider

v1.0.29

Custom Assets Facade Service

v1.0.7

Customization Service

v0.6.658

Dependency Resolver Service

v0.8.77

File Service

v1.1.288

File Upload Service

v1.1.226

Maven Facade Service

v0.8.109

Region Synchronizer Service

v0.2.163

Exchange API Metadata DB Migrations

v2.0.40

Exchange API Metadata Service

v1.1.1

Exchange Asset Reviews DB Migrations

v2.0.36

Exchange Asset Reviews Service

v0.8.435

Exchange Customization DB Migrations

2.0.31

Exchange Customization Service

0.6.658

Exchange Graph Node Service

0.2.664

Exchange Search DB Migrations

2.0.38

Exchange Search Service

v1.0.215

Exchange UI

v1.0.265

Exchange XAPI Service

v0.8.2848

Mocking Service

v2.77.0-17.pce

Runtime Manager

Application Manager DB Migration

v1.2.11

Application Manager Service

v2.33.51

Atlas Service

v1.9.121

Mule Agent

v2.6.6

Jafar Service

0.3.167

MCM Service

v2.17.347

Runtime Services

Anypoint Flash Platform Messages DB Migrations

1.1.43

Platform Messages

v2.2.43

Mule Message Processor

v3.2.40

## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/latest/install-checklist) - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/latest/install-workflow) - [Migrating Anypoint Platform PCE v3.X to v4.X](https://docs.mulesoft.com/private-cloud/latest/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Anypoint Platform PCE Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE): - Your organization’s operations, networking, and security teams must review the listed infrastructure prerequisites, including Amazon Web Services (if applicable), hardware, software, network, and NFS, and verify successful completion of all prerequisites listed in this topic. If needed, contact your MuleSoft representative for assistance. All prerequisites apply to all [Supported Environments](https://docs.mulesoft.com/private-cloud/3.2/prereq-platform#supported-environments). - Knowledge in the following areas is required: - Linux system commands. - Working in Linux with root access. Installation requires root access (permission to run the `sudo` command) on all servers in your environment. - Working with Terraform. This is only required if your environment includes AWS. - Remote connection mechanisms, such as SSL and SSH. > [!WARNING] You must contact MuleSoft Professional Services prior to installation. MuleSoft > professional services must be involved with your installation. > > Installing Anypoint Platform PCE requires specific versions of operating systems, NFS, and other > software. Most issues experienced with Anypoint Platform PCE occur because the required > environment was not set up prior to installing or upgrading. For example, the following situations > cause known problems: > > - Using your own AWS images when using AWS. > > - Using existing provisioning processes that install specific versions of RHEL, Docker, or other > software that Anypoint Platform PCE does not support. ## Environment Prerequisites
Table 1. Prerequisite Environment-Related Tasks
Task Description
## Hardware Prerequisites Refer to [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) for additional details and instructions for the following tasks:
Table 2. Prerequisite Hardware-Related Tasks
Task Description
## Software Prerequisites Refer to [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) for additional details and instructions for the following tasks:
Table 3. Prerequisite Software-Related Tasks
Task Description

Operating system prerequisites

Third party software prerequisites

MuleSoft software prerequisites

## Network Prerequisites Refer to [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) for additional details and instructions for the following tasks:
Table 4. Prerequisite Network-Related Tasks
Task Description
## NFS Prerequisites Refer to [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) for additional details and instructions for the following tasks:
Table 5. Prerequisite NFS-Related Tasks
Task Description
## Verify Environment Perform the steps in [Verify Environment](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) to ensure that your infrastructure meets Anypoint Platform PCE requirements. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Anypoint Platform PCE Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0 and later in the Kubernetes (K8s) cluster that you provision and manage. MuleSoft performed installations on popular K8s providers, which are listed in the following sections. The hardware and software requirements for installing Anypoint Platform PCE on your chosen provider are dictated by that provider, unless otherwise specified in this document. Before you begin, ensure your organization’s operations, networking, and security teams review the listed infrastructure prerequisites. These include the K8s cluster, hardware, software, networking, NFS storage, container registry, installer node, and storage requirements. > [!NOTE] You must contact MuleSoft Professional Services prior to installing Anypoint Platform PCE, > as they are involved with your installation. > > Installing Anypoint Platform PCE requires specific versions of operating systems, NFS, and other > software. Most issues experienced with Anypoint Platform PCE occur because the required > environment was not set up prior to installing or upgrading. ## Kubernetes Cluster Requirements These are the requirements: - Kubernetes (K8s) version 1.29 or 1.30. - Kubernetes providers: - Rancher Kubernetes Engine (RKE2) versions 1.29 and 1.30. - Red Hat OpenShift versions 4.15 and 4.16. - Amazon EKS Anywhere (EKS-A) 0.20 (K8s version 1.29, 1.30) - Cluster configurations: - Have a dedicated control plane node within the K8s cluster. Preferably, allocate 3 or 5 nodes for the K8s control planes. For example, only K8s management workload run on these nodes. - Have cluster worker nodes exclusively dedicated to Anypoint Platform PCE workloads. No Anypoint Platform PCE pods can run on the K8s control plane nodes. - A large cluster can have: - 3 K8s control plane nodes - 7 K8s worker nodes - A small cluster can have: - 3 K8s control plane nodes - 4 K8s worker nodes If you have Anypoint Platform PCE 3.2.x or earlier installed, ensure you disable SWAP on all Anypoint Platform PCE nodes. Anypoint Platform PCE relies heavily on Kubernetes, which doesn’t support the use of SWAP space. When SWAP is enabled, Kubernetes and the SWAP daemon attempt to manage the same memory space, leading to CPU overload as they compete. To avoid this issue, don’t use SWAP space on machines running Anypoint Platform PCE. > [!NOTE] Running Anypoint Platform PCE workloads on K8s control plane nodes can degrade both > Anypoint Platform PCE and cluster performance. ## Hardware Requirements These are the minimum requirements for the cluster nodes: - Dedicated control plane nodes: - Architecture x86\_64 - 4 vCPUS - Memory 8 GiB - Root volume 50GB - Worker nodes: - Architecture x86\_64 - 8 vCPUS - Memory 32 GiB - Root volume 50GB For more information, see the provider specific requirements: - [Rancher Kubernetes Engine (RKE2) - Hardware](https://docs.rke2.io/install/requirements#hardware) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/installation_overview/ocp-installation-overview) - [Amazon EKS Anywhere (EKS-A)](https://anywhere.eks.amazonaws.com/docs/getting-started/baremetal/bare-prereq/) ## Software Requirements These are the requirements: - Operating system: RHEL 8.10 For more information, see the provider specific requirements: - [Rancher Kubernetes Engine (RKE2) - Software](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/installation_overview/ocp-installation-overview) - [Amazon EKS Anywhere (EKS-A)](https://anywhere.eks.amazonaws.com/docs/getting-started/baremetal/bare-prereq/) ## Networking Requirements Use the following ports to communicate with Anypoint Platform PCE:
ProtocolPort / RangePurposeSourceDestination

TCP/UDP

53

Internal cluster DNS

localhost

localhost

TCP

2379, 2380, 4001, 7001

etcd distributed database

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

4242

Installer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

6443

Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9443

HAProxy to load balance Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9100

Prometheus node-exporter

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10248, 10250

Kubernetes Kubelet

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10249

Kubernetes kube-proxy

localhost

localhost

TCP

10251-10252

Kubernetes controller-manager and scheduler

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10257-10259

Kubernetes controller-manager and scheduler (secure)

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

30000-32767

Internal services port range

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

8472

Overlay VXLAN network

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

30080

HTTP public access

Load balancer

Anypoint Platform PCE nodes

TCP

30443

HTTPS public access

Load balancer

Anypoint Platform PCE nodes

TCP

30889

Mule Runtime Websocket

Load balancer

Anypoint Platform PCE nodes

TCP

30883

Mule Runtime Authentication Service (for certificate renewals)

Load balancer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes need to hit the platform API address on the load balancer via HTTPS. For more information, see the provider specific requirements: - [Rancher Kubernetes Engine (RKE2) - Networking](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift - Networking](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/networking/understanding-networking) - [Amazon EKS Anywhere (EKS-A) - Networking](https://anywhere.eks.amazonaws.com/docs/clustermgmt/networking/networking-and-security/) ## NFS Storage Requirements See [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/verify-nfs) ## Container Registry Requirements These are the requirements: - Must be secured over TLS. - Must be accessible from installer nodes and Kubernetes clusters. - Must be secured with either a username and password or Certificate based authentication. Ensure that the new or updated certificates are distributed to all nodes to facilitate uninterrupted TLS communication. ## Installer Node Requirements These are the requirements: - Operating system: RHEL, Ubuntu or any Linux based systems. - Storage required a minimum 100GB. - Access to Kubernetes cluster using kubeconfig.yaml - The cluster needs `cluster:admin` access, as Anypoint Platform PCE installer needs to create resources of different types. - The cluster needs `read/write` access in the default namespace to run Anypoint Platform PCE installer jobs and other supporting services. - Access and connectivity to private container registry - For OpenShift use SecurityContextConstraints for installing Anypoint Platform PCE. ## Storage Requirements These are the requirements: In Anypoint Platform PCE version 4.0 and later, all the persistent data is stored in the Persistent Volume Claims (PVCs). Kubernetes Storage Class commands the PVCs. Currently, there are two services that require PVCs: - SeaweedFS - S3 object store service - Stolon - PostgreSQL HA database The storage class for the previous services can be the same. Depending on the type of underlying infrastructure or storage solution you choose, the storage class configuration can vary. ## Platform Configuration To make Anypoint Platform PCE accessible, configure these specifications in the input config file before installing Anypoint Platform PCE: - Platform DNS - File System DNS - First User Account details - Org Name - Username - Email - Password - Platform certificate and certificate key with the Subject Alternative Name (SAN) of the platform DNS. ## See Also - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Anypoint Platform PCE Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.x and later in the Kubernetes (K8s) cluster that you provision and manage. MuleSoft performed installations on popular K8s providers, which are listed in the following sections. The hardware and software requirements for installing Anypoint Platform PCE on your chosen provider are dictated by that provider, unless otherwise specified in this document. Before you begin, make sure that your organization’s operations, networking, and security teams review the listed infrastructure prerequisites. These include the K8s cluster, hardware, software, networking, NFS storage, container registry, installer node, and storage requirements. > [!NOTE] Contact MuleSoft Professional Services prior to installing Anypoint Platform PCE, as they > are involved with your installation. > > Installing Anypoint Platform PCE requires specific versions of operating systems, NFS, and other > software. Most issues experienced with Anypoint Platform PCE occur because the required > environment was not set up prior to installing or upgrading. ## Kubernetes Cluster Requirements - Supported Kubernetes providers: - Amazon EKS Anywhere - Amazon EKS - Rancher Kubernetes Engine (RKE2) - Red Hat OpenShift - vSphere Kubernetes Service (VKS) - Cluster configurations: - Run Kubernetes control plane components on dedicated nodes. Preferably, allocate three nodes for the Kubernetes control plane. - Have cluster worker nodes exclusively dedicated to Anypoint Platform PCE workloads. Don’t run any Anypoint Platform PCE pods on the K8s control plane nodes. Check the prerequisites to make sure that PCE workloads can be scheduled and run on worker nodes. > [!NOTE] Running Anypoint Platform PCE workloads on K8s control plane nodes can degrade both > Anypoint Platform PCE and cluster performance. ## Hardware Requirements Anypoint Platform PCE runs as a Kubernetes application and these are the minimum required hardware for all services to be scheduled and running: - 32 vCPUs - 128 GiB Memory - x86\_64 Architecture You can provision and distribute these resources across a minimum of four worker nodes. PCE requires at least three worker nodes as hosts for data services that are deployed and configured in a cluster setup with three replicas. Make sure these three replicas run on three separate nodes to honor the anti-affinity rules. ### Sample Worker Node Configurations Anypoint Platform PCE offers flexibility in how you configure your Kubernetes worker nodes. You can use different numbers of worker nodes (for example, 4, 5, 8, or 10 nodes) with varying resource allocations per node. The key requirement is that your cluster’s total resources meet or exceed the minimum requirements of 32 vCPUs and 128 GiB memory. This table shows sample worker node configurations that MuleSoft’s performance teams tested and validated:
Number of Worker NodesNode CompositionTotal Resources

4

8 vCPU and 32 GiB memory

32 vCPU and 128 GiB memory

10

4 vCPU and 16 GiB memory

40 vCPU and 160 GiB memory

> [!NOTE] These configurations are reference examples to help you understand how Anypoint Platform > PCE can work with different worker node configurations. They are not strict requirements. You can > customize your cluster configuration based on your infrastructure and operational needs, as long > as the total resource requirements are met. ### Load Balancer Configuration You can configure your load balancer to use any method for distributing client requests. Use a a round robin strategy for better results in most contexts. Make sure the load balancer is accessible through an IP address that is available to all machines on your network. Make sure your load balancer routes these TCP ports:
Table 1. TCP Port Routing Requirements
Load Balancer PortInstance PortInternal Usage

80

30080

HTTP redirects to HTTPs

443

30443

HTTPS port

8083

30883

HTTPS port for Mule runtimes to authenticate and renew certificates

8889

30889

WebSocket port for Mule runtimes to connect

In each case, configure your load balancer to listen on the load balancer port and redirect incoming requests to the instance port. Your Anypoint Platform PCE installation includes an internal NGINX server that listens on each of the configured instance ports, and then performs the action listed in the Internal Usage column. Make sure your load balancer polls the address `HTTP:10256/healthz` to run a health check on your platform servers and confirm that they are accessible. > [!NOTE] Don’t configure SSL certificates in your load balancer. The platform handles the TLS > termination with the certificates configured using Access Management. See > [Configure Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow). For more information, see the provider specific requirements: - [Amazon EKS](https://docs.aws.amazon.com/eks/latest/userguide/create-cluster.html) AWS EKS is only tested with default networking add-ons. - [Amazon EKS Anywhere (EKS-A)](https://anywhere.eks.amazonaws.com/docs/getting-started/baremetal/bare-prereq/) - [Rancher Kubernetes Engine (RKE2) - Hardware](https://docs.rke2.io/install/requirements#hardware) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/installation_overview/ocp-installation-overview) - [vSphere Kubernetes Service (VKS)](https://techdocs.broadcom.com/us/en/vmware-cis/vcf/vsphere-supervisor-services-and-standalone-components/latest/managing-vsphere-kubernetes-service/running-tkg-service-clusters/tkg-service-components.html) ## Software Requirements These are the requirements: - Operating system: RHEL 8.10, 9.4, 9.6 For more information, see the provider specific requirements: - [Amazon EKS](https://docs.aws.amazon.com/eks/latest/userguide/create-cluster.html) - [Amazon EKS Anywhere (EKS-A)](https://anywhere.eks.amazonaws.com/docs/getting-started/baremetal/bare-prereq/) - [Rancher Kubernetes Engine (RKE2) - Software](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/installation_overview/ocp-installation-overview) ## Networking Requirements Use the following ports to communicate with Anypoint Platform PCE:
ProtocolPort / RangePurposeSourceDestination

TCP/UDP

53

Internal cluster DNS

localhost

localhost

TCP

2379, 2380, 4001, 7001

etcd distributed database

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

4242

Installer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

6443

Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9443

HAProxy to load balance Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9100

Prometheus node-exporter

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10248, 10250

Kubernetes Kubelet

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10249

Kubernetes kube-proxy

localhost

localhost

TCP

10251-10252

Kubernetes controller-manager and scheduler

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10257-10259

Kubernetes controller-manager and scheduler (secure)

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

30000-32767

Internal services port range

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

8472

Overlay VXLAN network

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

30080

HTTP public access

Load balancer

Anypoint Platform PCE nodes

TCP

30443

HTTPS public access

Load balancer

Anypoint Platform PCE nodes

TCP

30889

Mule Runtime Websocket

Load balancer

Anypoint Platform PCE nodes

TCP

30883

Mule Runtime Authentication Service (for certificate renewals)

Load balancer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes need to hit the platform API address on the load balancer via HTTPS. For more information, see the provider specific requirements: - [Amazon EKS - Networking](https://docs.aws.amazon.com/eks/latest/userguide/eks-networking.html) - [Amazon EKS Anywhere (EKS-A) - Networking](https://anywhere.eks.amazonaws.com/docs/clustermgmt/networking/networking-and-security/) - [Rancher Kubernetes Engine (RKE2) - Networking](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift - Networking](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/networking/understanding-networking) ## NFS Storage Requirements See [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/verify-nfs) ## Container Registry Requirements These are the requirements: - Must be secured over TLS. - Must be accessible from installer nodes and Kubernetes clusters. - Must be secured with either a username and password, or Certificate-based authentication. Make sure that the new or updated certificates are distributed to all nodes to facilitate uninterrupted TLS communication. ## Installer Node Requirements These are the requirements: - Operating system: RHEL, Ubuntu, or any Linux-based systems. - Storage required a minimum 100GB. - Access to Kubernetes cluster using kubeconfig.yaml - The cluster needs `cluster:admin` access, as Anypoint Platform PCE installer needs to create resources of different types. - The cluster needs `read/write` access in the pce-core namespace to run Anypoint Platform PCE installer jobs and other supporting services. - Access and connectivity to private container registry - For OpenShift use SecurityContextConstraints for installing Anypoint Platform PCE. Before executing the `pcectl` binary for Anypoint Platform PCE 4.1, make sure that you follow these requirements: - Use a Linux based operating system - Use glibc (GNU C library) version 2.28 or higher To check the glibc version, run this command: ldd --version - Use Python version 3.12 or newer To check your Python version, run this command: python --version - Use gpg (GnuPG) version 2.1 or higher To check your gpg version, run this command: gpg --version ## Storage Requirements These are the requirements: In Anypoint Platform PCE version 4.1 and later, all the persistent data is stored in the Persistent Volume Claims (PVCs). Kubernetes Storage Class commands the PVCs. These services use PVCs with minimum and default storage size: - SeaweedFS - S3 Object Store service - Total 120Gi - Stolon - PostgreSQL HA database - Total 100 Gi - Prometheus - TSDB for k8s metrics (optional) - Default 50Gi The storage class for the previous services can be the same. Depending on the type of underlying infrastructure or storage solution you select, the storage class configuration can vary. You can update the PVC storage size during the installation by updating the `input.yaml` file. See these sections to update the storage size: ```json # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "20Gi" # Optional. Defaults to 10Gi volumeSize: "200Gi" # Optional. Defaults to 100Gi volumeIndexSize: "20Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "300Gi" # Optional. Defaults to 100Gi ``` In the previous example, you can see the storage for SeaweedFS that is Object Store has increased to 240GiB and stolon has increased to 300 Gi. Similarly, to set the storage size for monitoring stack, use this configuration: ```json # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStack: enabled: True # Defaults to True. Must be set to False for Openshift clusters. values: | # Optional. If not provided, the default values will be used. prometheus: prometheusSpec: storageSpec: volumeClaimTemplate: spec: storageClassName: 'local-path', accessModes: ["ReadWriteOnce"] resources: requests: storage: 50Gi ``` ## Platform Configuration To make Anypoint Platform PCE accessible, configure these specifications in the input configuration file before installing Anypoint Platform PCE: - Platform DNS - File System DNS - First User Account details - Org Name - Username - Email - Password - Platform certificate and certificate key with the Subject Alternative Name (SAN) of the platform DNS. ## See Also - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Anypoint Platform PCE Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.x and later in the Kubernetes (K8s) cluster that you provision and manage. MuleSoft performed installations on popular K8s providers, which are listed in the following sections. The hardware and software requirements for installing Anypoint Platform PCE on your chosen provider are dictated by that provider, unless otherwise specified in this document. Before you begin, make sure that your organization’s operations, networking, and security teams review the listed infrastructure prerequisites. These include the K8s cluster, hardware, software, networking, NFS storage, container registry, installer node, and storage requirements. > [!NOTE] Contact MuleSoft Professional Services prior to installing Anypoint Platform PCE, as they > are involved with your installation. > > Installing Anypoint Platform PCE requires specific versions of operating systems, NFS, and other > software. Most issues experienced with Anypoint Platform PCE occur because the required > environment was not set up prior to installing or upgrading. ## Kubernetes Cluster Requirements - Supported Kubernetes providers: - Amazon EKS - Rancher Kubernetes Engine (RKE2) - Red Hat OpenShift - vSphere Kubernetes Service (VKS) - Cluster configurations: - Run Kubernetes control plane components on dedicated nodes. Preferably, allocate three nodes for the Kubernetes control plane. - Have cluster worker nodes exclusively dedicated to Anypoint Platform PCE workloads. Don’t run any Anypoint Platform PCE pods on the K8s control plane nodes. Check the prerequisites to make sure that PCE workloads can be scheduled and run on worker nodes. > [!NOTE] Running Anypoint Platform PCE workloads on K8s control plane nodes can degrade both > Anypoint Platform PCE and cluster performance. ## Hardware Requirements Anypoint Platform PCE runs as a Kubernetes application and these are the minimum required hardware for all services to be scheduled and running: - 32 vCPUs - 128 GiB Memory - x86\_64 Architecture You can provision and distribute these resources across a minimum of four worker nodes. PCE requires at least three worker nodes as hosts for data services that are deployed and configured in a cluster setup with three replicas. Make sure these three replicas run on three separate nodes to honor the anti-affinity rules. ### Sample Worker Node Configurations Anypoint Platform PCE offers flexibility in how you configure your Kubernetes worker nodes. You can use different numbers of worker nodes (for example, 4, 5, 8, or 10 nodes) with varying resource allocations per node. The key requirement is that your cluster’s total resources meet or exceed the minimum requirements of 32 vCPUs and 128 GiB memory. This table shows sample worker node configurations that MuleSoft’s performance teams tested and validated:
Number of Worker NodesNode CompositionTotal Resources

4

8 vCPU and 32 GiB memory

32 vCPU and 128 GiB memory

10

4 vCPU and 16 GiB memory

40 vCPU and 160 GiB memory

> [!NOTE] These configurations are reference examples to help you understand how Anypoint Platform > PCE can work with different worker node configurations. They are not strict requirements. You can > customize your cluster configuration based on your infrastructure and operational needs, as long > as the total resource requirements are met. ### Load Balancer Configuration You can configure your load balancer to use any method for distributing client requests. Use a a round robin strategy for better results in most contexts. Make sure the load balancer is accessible through an IP address that is available to all machines on your network. Make sure your load balancer routes these TCP ports:
Table 1. TCP Port Routing Requirements
Load Balancer PortInstance PortInternal Usage

80

30080

HTTP redirects to HTTPs

443

30443

HTTPS port

8083

30883

HTTPS port for Mule runtimes to authenticate and renew certificates

8889

30889

WebSocket port for Mule runtimes to connect

In each case, configure your load balancer to listen on the load balancer port and redirect incoming requests to the instance port. Your Anypoint Platform PCE installation includes an internal NGINX server that listens on each of the configured instance ports, and then performs the action listed in the Internal Usage column. Make sure your load balancer polls the address `HTTP:10256/healthz` to run a health check on your platform servers and confirm that they are accessible. > [!NOTE] Don’t configure SSL certificates in your load balancer. The platform handles the TLS > termination with the certificates configured using Access Management. See > [Configure Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow). For more information, see the provider specific requirements: - [Amazon EKS](https://docs.aws.amazon.com/eks/latest/userguide/create-cluster.html) AWS EKS is only tested with default networking add-ons. - [Rancher Kubernetes Engine (RKE2) - Hardware](https://docs.rke2.io/install/requirements#hardware) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/installation_overview/ocp-installation-overview) - [vSphere Kubernetes Service (VKS)](https://techdocs.broadcom.com/us/en/vmware-cis/vcf/vsphere-supervisor-services-and-standalone-components/latest/managing-vsphere-kubernetes-service/running-tkg-service-clusters/tkg-service-components.html) ## Software Requirements These are the requirements: - Operating system: RHEL 8.10, 9.4, 9.6 For more information, see the provider specific requirements: - [Amazon EKS](https://docs.aws.amazon.com/eks/latest/userguide/create-cluster.html) - [Rancher Kubernetes Engine (RKE2) - Software](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/installation_overview/ocp-installation-overview) ## Networking Requirements Use these ports to communicate with Anypoint Platform PCE:
ProtocolPort / RangePurposeSourceDestination

TCP/UDP

53

Internal cluster DNS

localhost

localhost

TCP

2379, 2380, 4001, 7001

etcd distributed database

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

4242

Installer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

6443

Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9443

HAProxy to load balance Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9100

Prometheus node-exporter

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10248, 10250

Kubernetes Kubelet

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10249

Kubernetes kube-proxy

localhost

localhost

TCP

10251-10252

Kubernetes controller-manager and scheduler

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10257-10259

Kubernetes controller-manager and scheduler (secure)

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

30000-32767

Internal services port range

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

8472

Overlay VXLAN network

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

30080

HTTP public access

Load balancer

Anypoint Platform PCE nodes

TCP

30443

HTTPS public access

Load balancer

Anypoint Platform PCE nodes

TCP

30889

Mule Runtime Websocket

Load balancer

Anypoint Platform PCE nodes

TCP

30883

Mule Runtime Authentication Service (for certificate renewals)

Load balancer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes need to hit the platform API address on the load balancer via HTTPS. For more information, see the provider specific requirements: - [Amazon EKS - Networking](https://docs.aws.amazon.com/eks/latest/userguide/eks-networking.html) - [Rancher Kubernetes Engine (RKE2) - Networking](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift - Networking](https://docs.redhat.com/en/documentation/openshift_container_platform/4.16/html/networking/understanding-networking) ## NFS Storage Requirements A pre-installed Container Storage Interface (CSI) driver in the Kubernetes cluster is a mandatory requirement for Anypoint Platform PCE 4.2.0. Anypoint Platform PCE uses the CSI driver to mount NFS volumes into pods. For detailed configuration instructions, including EFS-specific volume handle setup and NFS mount option overrides, see [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/4.2/verify-nfs). ## Container Registry Requirements These are the requirements: - Must be secured over TLS. - Must be accessible from installer nodes and Kubernetes clusters. - Must be secured with either a username and password, or Certificate-based authentication. Make sure that the new or updated certificates are distributed to all nodes to facilitate uninterrupted TLS communication. ## Installer Node Requirements These are the requirements: - Operating system: RHEL, Ubuntu, or any Linux-based systems. - Storage required a minimum 100GB. - Access to Kubernetes cluster using kubeconfig.yaml - The cluster needs `cluster:admin` access, as Anypoint Platform PCE installer needs to create resources of different types. - The cluster needs `read/write` access in the pce-core namespace to run Anypoint Platform PCE installer jobs and other supporting services. - Access and connectivity to private container registry - For OpenShift use SecurityContextConstraints for installing Anypoint Platform PCE. Before executing the `pcectl` binary for Anypoint Platform PCE 4.2, make sure that you follow these requirements: - Use a Linux based operating system - Use glibc (GNU C library) version 2.28 or higher To check the glibc version, run this command: ldd --version - Use Python version 3.12 or newer To check your Python version, run this command: python --version - Use gpg (GnuPG) version 2.1 or higher To check your gpg version, run this command: gpg --version ## Storage Requirements These are the requirements: In Anypoint Platform PCE version 4.2 and later, all the persistent data is stored in the Persistent Volume Claims (PVCs). Kubernetes Storage Class commands the PVCs. These services use PVCs with minimum and default storage size: - SeaweedFS - S3 Object Store service - Total 120Gi - Stolon - PostgreSQL HA database - Total 100 Gi - Prometheus - TSDB for k8s metrics (optional) - Default 50Gi The storage class for the previous services can be the same. Depending on the type of underlying infrastructure or storage solution you select, the storage class configuration can vary. You can update the PVC storage size during the installation by updating the `input.yaml` file. See these sections to update the storage size: ```json # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "20Gi" # Optional. Defaults to 10Gi volumeSize: "200Gi" # Optional. Defaults to 100Gi volumeIndexSize: "20Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "300Gi" # Optional. Defaults to 100Gi ``` In the previous example, you can see the storage for SeaweedFS that is Object Store has increased to 240GiB and stolon has increased to 300 Gi. Similarly, to set the storage size for monitoring stack, use this configuration: ```json # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStack: enabled: True # Defaults to True. Must be set to False for Openshift clusters. values: | # Optional. If not provided, the default values will be used. prometheus: prometheusSpec: storageSpec: volumeClaimTemplate: spec: storageClassName: 'local-path', accessModes: ["ReadWriteOnce"] resources: requests: storage: 50Gi ``` ## Platform Configuration To make Anypoint Platform PCE accessible, configure these specifications in the input configuration file before installing Anypoint Platform PCE: - Platform DNS - File System DNS - First User Account details - Org Name - Username - Email - Password - Platform certificate and certificate key with the Subject Alternative Name (SAN) of the platform DNS. ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Anypoint Platform PCE Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.x and later in the Kubernetes (K8s) cluster that you provision and manage. MuleSoft performed installations on popular K8s providers, which are listed in the following sections. The hardware and software requirements for installing Anypoint Platform PCE on your chosen provider are dictated by that provider, unless otherwise specified in this document. Before you begin, make sure that your organization’s operations, networking, and security teams review the listed infrastructure prerequisites. These include the K8s cluster, hardware, software, networking, shared filesystem, container registry, installer node, and storage requirements. > [!NOTE] Contact MuleSoft Professional Services prior to installing Anypoint Platform PCE, as they > are involved with your installation. > > Installing Anypoint Platform PCE requires specific versions of operating systems, shared > filesystem components, and other software. Most issues experienced with Anypoint Platform PCE > occur because the required environment was not set up prior to installing or upgrading. ## Kubernetes Cluster Requirements - Supported Kubernetes providers: - Amazon EKS - Rancher Kubernetes Engine (RKE2) - Red Hat OpenShift - vSphere Kubernetes Service (VKS) - Cluster configurations: - Run Kubernetes control plane components on dedicated nodes. Preferably, allocate three nodes for the Kubernetes control plane. - Have cluster worker nodes exclusively dedicated to Anypoint Platform PCE workloads. Don’t run any Anypoint Platform PCE pods on the K8s control plane nodes. Check the prerequisites to make sure that PCE workloads can be scheduled and run on worker nodes. > [!NOTE] Running Anypoint Platform PCE workloads on K8s control plane nodes can degrade both > Anypoint Platform PCE and cluster performance. ## Hardware Requirements Anypoint Platform PCE runs as a Kubernetes application and these are the minimum required hardware for all services to be scheduled and running: - 32 vCPUs - 128 GiB Memory - x86\_64 Architecture You can provision and distribute these resources across a minimum of four worker nodes. PCE requires at least three worker nodes as hosts for data services that are deployed and configured in a cluster setup with three replicas. Make sure these three replicas run on three separate nodes to honor the anti-affinity rules. ### Sample Worker Node Configurations Anypoint Platform PCE offers flexibility in how you configure your Kubernetes worker nodes. You can use different numbers of worker nodes (for example, 4, 5, 8, or 10 nodes) with varying resource allocations per node. The key requirement is that your cluster’s total resources meet or exceed the minimum requirements of 32 vCPUs and 128 GiB memory. This table shows sample worker node configurations that MuleSoft’s performance teams tested and validated:
Number of Worker NodesNode CompositionTotal Resources

4

8 vCPU and 32 GiB memory

32 vCPU and 128 GiB memory

10

4 vCPU and 16 GiB memory

40 vCPU and 160 GiB memory

> [!NOTE] These configurations are reference examples to help you understand how Anypoint Platform > PCE can work with different worker node configurations. They are not strict requirements. You can > customize your cluster configuration based on your infrastructure and operational needs, as long > as the total resource requirements are met. ### Load Balancer Configuration You can configure your load balancer to use any method for distributing client requests. Use a a round robin strategy for better results in most contexts. Make sure the load balancer is accessible through an IP address that is available to all machines on your network. Make sure your load balancer routes these TCP ports:
Table 1. TCP Port Routing Requirements
Load Balancer PortInstance PortInternal Usage

80

30080

HTTP redirects to HTTPs

443

30443

HTTPS port

8083

30883

HTTPS port for Mule runtimes to authenticate and renew certificates

8889

30889

WebSocket port for Mule runtimes to connect

In each case, configure your load balancer to listen on the load balancer port and redirect incoming requests to the instance port. Your Anypoint Platform PCE installation includes an internal NGINX server that listens on each of the configured instance ports, and then performs the action listed in the Internal Usage column. Make sure your load balancer polls the address `HTTP:10256/healthz` to run a health check on your platform servers and confirm that they are accessible. > [!NOTE] Don’t configure SSL certificates in your load balancer. The platform handles the TLS > termination with the certificates configured using Access Management. See > [Configure Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow). For more information, see the provider specific requirements: - [Amazon EKS](https://docs.aws.amazon.com/eks/latest/userguide/create-cluster.html) AWS EKS is only tested with default networking add-ons. - [Rancher Kubernetes Engine (RKE2) - Hardware](https://docs.rke2.io/install/requirements#hardware) - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.21/html/installation_overview/ocp-installation-overview) - [vSphere Kubernetes Service (VKS)](https://techdocs.broadcom.com/us/en/vmware-cis/vcf/vsphere-supervisor-services-and-standalone-components/latest/managing-vsphere-kubernetes-service/running-tkg-service-clusters/tkg-service-components.html) ## Software Requirements Provider specific node OS requirements: - [Amazon EKS](https://docs.aws.amazon.com/eks/latest/userguide/create-cluster.html) - [Rancher Kubernetes Engine (RKE2) - Software](https://docs.rke2.io/install/requirements#operating-systems) Validated against RHEL 9.6 and Kubernetes 1.34. - [Red Hat OpenShift](https://docs.redhat.com/en/documentation/openshift_container_platform/4.21/html/installation_overview/ocp-installation-overview) Validated against Red Hat Enterprise Linux CoreOS (RHCOS) 9.6 and OpenShift 4.21. ## Networking Requirements Use these ports to communicate with Anypoint Platform PCE:
ProtocolPort / RangePurposeSourceDestination

TCP

30080

HTTP public access

Load balancer

Anypoint Platform PCE nodes

TCP

30443

HTTPS public access

Load balancer

Anypoint Platform PCE nodes

TCP

30889

Mule Runtime Websocket

Load balancer

Anypoint Platform PCE nodes

TCP

30883

Mule Runtime Authentication Service (for certificate renewals)

Load balancer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes need to hit the platform API address on the load balancer via HTTPS. For more information, see the provider specific requirements: - [Amazon EKS - Networking](https://docs.aws.amazon.com/eks/latest/userguide/eks-networking.html) - [Rancher Kubernetes Engine (RKE2) - Networking](https://docs.rke2.io/install/requirements#operating-systems) - [Red Hat OpenShift - Networking](https://docs.redhat.com/en/documentation/openshift_container_platform/4.21/html/networking_overview/understanding-networking) ## Shared Filesystem Requirements Anypoint Platform PCE requires a `ReadWriteMany` (RWX) shared filesystem. Starting with Anypoint Platform PCE 4.3.0, you can provide this filesystem using either of the following storage modes: - **NFS mode** — Bind the platform to an existing NFS server through a pre-installed Container Storage Interface (CSI) driver in the Kubernetes cluster. This is the only supported mode in versions before 4.3.0. - **StorageClass mode** — Provide a single RWX-capable Kubernetes `StorageClass` (for example, backed by CephFS) and let Anypoint Platform PCE dynamically provision the shared volumes. This mode removes the requirement to operate an external NFS server. In StorageClass mode, Anypoint Platform PCE requests these persistent volumes: - Design Center: 2 PVCs, 300 GiB each - Backup and restore: 1 PVC, 1 TiB The two modes are mutually exclusive. You cannot change the storage mode on an existing cluster. The installer rejects an in-place upgrade from NFS mode to StorageClass mode so that data on the previous volumes is not orphaned. Switching the NFS CSI driver during an upgrade does not require additional storage changes. For detailed configuration instructions, including CSI driver setup, EFS-specific volume handle setup, NFS mount option overrides, and `StorageClass` requirements, see [Shared Filesystem Prerequisites](https://docs.mulesoft.com/private-cloud/latest/verify-nfs). ## Container Registry Requirements These are the requirements: - Must be secured over TLS. - Must be accessible from installer nodes and Kubernetes clusters. - Must be secured with either a username and password, or Certificate-based authentication. Make sure that the new or updated certificates are distributed to all nodes to facilitate uninterrupted TLS communication. ## Installer Node Requirements These are the requirements: - Operating system: RHEL, Ubuntu, or any Linux-based systems. - Storage required a minimum 100GB. - Access to Kubernetes cluster using kubeconfig.yaml - The cluster needs `cluster:admin` access, as Anypoint Platform PCE installer needs to create resources of different types. - The cluster needs `read/write` access in the pce-core namespace to run Anypoint Platform PCE installer jobs and other supporting services. - Access and connectivity to private container registry - For OpenShift use SecurityContextConstraints for installing Anypoint Platform PCE. Before executing the `pcectl` binary for Anypoint Platform PCE 4.2, make sure that you follow these requirements: - Use a Linux based operating system - Use glibc (GNU C library) version 2.28 or higher To check the glibc version, run this command: ldd --version - Use Python version 3.12 or newer To check your Python version, run this command: python --version - Use gpg (GnuPG) version 2.1 or higher To check your gpg version, run this command: gpg --version ## Storage Requirements These are the requirements: In Anypoint Platform PCE version 4.2 and later, all the persistent data is stored in the Persistent Volume Claims (PVCs). Kubernetes Storage Class commands the PVCs. These services use PVCs with minimum and default storage size: - SeaweedFS - S3 Object Store service - Total 120Gi - Stolon - PostgreSQL HA database - Total 100 Gi - Prometheus - TSDB for k8s metrics (optional) - Default 50Gi The storage class for the previous services can be the same. Depending on the type of underlying infrastructure or storage solution you select, the storage class configuration can vary. You can update the PVC storage size during the installation by updating the `input.yaml` file. See these sections to update the storage size: ```json # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "20Gi" # Optional. Defaults to 10Gi volumeSize: "200Gi" # Optional. Defaults to 100Gi volumeIndexSize: "20Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "300Gi" # Optional. Defaults to 100Gi ``` In the previous example, you can see the storage for SeaweedFS that is Object Store has increased to 240GiB and stolon has increased to 300 Gi. Similarly, to set the storage size for monitoring stack, use this configuration: ```json # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStack: enabled: True # Defaults to True. Must be set to False for Openshift clusters. values: | # Optional. If not provided, the default values will be used. prometheus: prometheusSpec: storageSpec: volumeClaimTemplate: spec: storageClassName: 'local-path', accessModes: ["ReadWriteOnce"] resources: requests: storage: 50Gi ``` ## Platform Configuration To make Anypoint Platform PCE accessible, configure these specifications in the input configuration file before installing Anypoint Platform PCE: - Platform DNS - File System DNS - First User Account details - Org Name - Username - Email - Password - Platform certificate and certificate key with the Subject Alternative Name (SAN) of the platform DNS. ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/latest/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Anypoint Platform Private Cloud Edition Release Notes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt - [Anypoint Platform Private Cloud Edition 4.3.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.3.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.3-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.2-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.0.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.0.0-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.5 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.5-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.4 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.4-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.3-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.2-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.0-release-notes) ## See Also - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Anypoint Platform Private Cloud Edition Release Notes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt - [Anypoint Platform Private Cloud Edition 4.3.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.3.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.3-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.2-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.0.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.0.0-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.5 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.5-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.4 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.4-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.3-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.2-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.0-release-notes) ## See Also - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Anypoint Platform Private Cloud Edition Release Notes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt - [Anypoint Platform Private Cloud Edition 4.3.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.3.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.3-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.2-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.0.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.0.0-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.5 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.5-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.4 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.4-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.3-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.2-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.0-release-notes) ## See Also - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Anypoint Platform Private Cloud Edition Release Notes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt - [Anypoint Platform Private Cloud Edition 4.3.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.3.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.3-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.2-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.0.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.0.0-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.5 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.5-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.4 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.4-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.3-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.2-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.0-release-notes) ## See Also - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Anypoint Platform Private Cloud Edition Release Notes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt - [Anypoint Platform Private Cloud Edition 4.3.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.3.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 4.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.2.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.3-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.2-release-notes) - [Anypoint Platform Private Cloud Edition 4.1.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.1.0-release-notes) - [Anypoint Platform Private Cloud Edition 4.0.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-4.0.0-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.5 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.5-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.4 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.4-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.3 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.3-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.2 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.2-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.1 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.1-release-notes) - [Anypoint Platform Private Cloud Edition 3.2.0 Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-3.2.0-release-notes) ## See Also - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # AWS Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you are using Amazon Web Services (AWS), you must create the resources required to install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) on AWS. Anypoint Platform PCE supports 4-node and 7-node configurations in a production environment on AWS. You do not need this information if you are starting from bare metal servers. > [!IMPORTANT] Your infrastructure team, or someone with AWS administrator level knowledge, must > perform the following tasks. If needed, contact your MuleSoft representative for help. ## AWS Account Permissions and Resources To install Anypoint Platform PCE on AWS: - Your AWS account must have AWS keys with `EC2FullAccess` and `S3FullAccess` permissions. - When you create your AWS environment, the following resources are created:
Table 1. AWS-Created Resources
AWS ResourceNumber Required (4-node)Number Required (7-node)Anypoint Monitoring (Optional)

m5.8xlarge

0

0

3

m5.2xlarge

4

7

0

root disk @ 500 iops

4

7

3

EBS volumes @ 1500 iops

8

14

12

EBS volume @ 3000 iops

8

14

0

Amazon ELB

1

1

0

t2.medium

1

1

0

## Run the AWS Provisioner MuleSoft provides a Docker image that you can use to provision the resources for your AWS account. You can also run additional custom shell scripts on the provisioned instances as shown in the Custom provisioning scripts section. 1. Create an initial instance (t2.small) on your AWS account on any VPC with internet access. This instance runs the actual provisioning of the cluster. You must use an AMI that has Docker installed by default or you must manually install Docker after creating the AWS instance. You can also run provisioner remotely from any other machine with Docker and internet access. 2. Download the Private Cloud Provisioner (PCP) Docker image from: `[https://anypoint-anywhere.s3.amazonaws.com/pcp/private-cloud-provisioner-3.1.14.tar.gz?Signature=jB4UiQgzsSSE3iyaPWYNJ67dCyM%3D&AWSAccessKeyId=AKIAI6LBJNQFR2V3CQWQ&Expires=1801445392](https://anypoint-anywhere.s3.amazonaws.com/pcp/private-cloud-provisioner-3.1.14.tar.gz?Signature=jB4UiQgzsSSE3iyaPWYNJ67dCyM%3D&AWSAccessKeyId=AKIAI6LBJNQFR2V3CQWQ&Expires=1801445392)` 3. Copy the provisioner Docker image to the instance via SCP: scp -i .pem ~/Downloads/private-cloud-provisioner-3.1.14.tar.gz ec2-user@W.X.Y.Z:/home/ec2-user 4. SSH into the instance: ssh -i 'anypoint.pem' ec2-user@W.X.Y.Z 5. Create the variable file (pce.env) with the environment details using the following values:
Table 2. Required Environment Variables
NameDescription

AWS_ACCESS_KEY_ID

Specifies the AWS access ID Terraform uses to connect to your AWS account.

AWS_SECRET_ACCESS_KEY

Specifies the AWS access key Terraform uses to connect to your AWS account.

AWS_SESSION_TOKEN

Specifies the AWS temporary session token, if one exists.

AWS_KEY_NAME

Specifies the AWS SSH key. Do not include the .pem extension.

AWS_REGION

Specifies the AWS region where Terraform creates the cluster, for example, us-east-2.

TF_VAR_ssh_user=<SSH_USER>

Specifies the ssh user, for example, ec2-user, centos, etc.

TF_VAR_trusted_cidrs_list

List of trusted CIDRs that can be accessed from the environment. Use [0.0.0.0/0] to open the cluster to the internet.

TELEKUBE_CLUSTER_NAME

Specifies the name of the cluster and the corresponding AWS resources.

TELEKUBE_NODE_PROFILES

Must be set to node.

TELEKUBE_NODE_PROFILE_COUNT_node

Specifies the number of nodes in your cluster. Possible values are 4 and 7.

TELEKUBE_NODE_PROFILE_INSTANCE_TYPE_node

Must be m5.2xlarge.

TELEKUBE_NODE_PROFILE_COUNT_node_amv

Specifies the number of nodes for the AMV addon in your cluster. Must be 3 if planning to install the AMV addon. Omit otherwise.

TELEKUBE_NODE_PROFILE_INSTANCE_TYPE_node_amv

Must be m5.8xlarge if planning to install the AMV addon. Omit otherwise.

TF_VAR_high_performance_disks

Must be set to true for production environments.

You can also include the following set of optional environment variables:
Table 3. Optional Environment Variables
NameDescription

TF_VAR_ami_name

Specifies the AMI name to be used for the instances. Use the AMI name, NOT the AMI ID. Ensure that the AMI does not provision any additional volumes. If you do not set this variable, the provisioner uses the following AMI by default: RHEL-8.3.0_HVM-20201031-x86_64-0-Hourly2-GP2

TF_VAR_monitoring=<true or false>

If true, creates basic instance alarms in Cloudwatch.

TF_VAR_use_bastion=<true or false>

If true, creates a small instance (using an ASG) in a public subnet as a jumpbox and associates a public IP address, and launches the cluster instances in the private subnets.

TF_VAR_internal=<true or false>

If true, the cluster instances are launched in private subnets and do not have public IPs associated with them. Also, the provisioned load balancer is internal only.

AWS_VPC_ID=<VPC_ID>

Specifies an existing VPC that should already have an internet gateway attached to it. Subnets are still provisioned within the existing VPC.

AWS_VPC_CIDR=<VPC_CIDR>

Enables you to pass a preferred CIDR for the new VPC to be provisioned. (Requires you to not pass a AWS_VPC_ID).

TF_VAR_aws_selected_subnets_ids=<SUBNETS_IDS>

Enables you to pass a list of subnets IDs on which the instances are launched. If you reuse your own subnets, you need to set up Route Tables, NAT Gateways, and an Internet Gateway, for example, ["subnet-0a17317984065f98f", "subnet-0600b4befb27c7949", "subnet-02103e0c935eff75a"].

TF_VAR_subnets=<PRIVATE_SUBNETS_CIDRs>

Enables you to optionally pass the preferred CIDRs blocks for the NEW subnets to be provisioned. You must pass the exact minimum number between the number of AZs of the region and the number of nodes you selected for your cluster, for example, '["10.1.0.0/24", "10.1.1.0/24", "10.1.2.0/24"]'.

TF_VAR_public_subnets=<PUBLIC_SUBNETS_CIDRs>

Enables you to optionally pass the preferred CIDRs blocks for the NEW subnets to be provisioned. You must pass the exact minimum number between the number of AZs of the region and the number of nodes you selected for your cluster, for example, '["10.1.3.0/24", "10.1.4.0/24", "10.1.5.0/24"]'.

TF_VAR_role_tag_value=<AWS_TAG_VALUE_FOR_ROLE_LABEL>

Enables you to specify a ROLE tag value to be applied to all AWS resources.

6. Load the provisioner Docker image into the local Docker registry: docker load -i private-cloud-provisioner-4.0.13.tar.gz 7. Perform a dry-run test: docker run --rm --env-file pce.env artifacts.msap.io/mulesoft/core-paas-private-cloud-provisioner:v4.0.13 dry-run 8. Run the provisioner: docker run --rm --env-file pce.env artifacts.msap.io/mulesoft/core-paas-private-cloud-provisioner:v4.0.13 cluster-provision After the provisioner runs successfully, it displays information about your environment including IP addresses and DNS name of the load balancer. 9. Verify that the provisioning script ran successfully by checking the existence of /var/lib/bootstrap\_complete on the instances ## Custom Provisioning Scripts (Optional) You can have your own shell scripts to run on the provisioned instances before and/or after the PCE provisioner scripts. To do so, place your shell scripts with .sh extension inside a folder named `pre-user-data` and/or `post-user-data`. Include the following volume mounts on the docker run command: docker run --rm -v $(pwd)/pre-user-data:/usr/local/bin/provisioner/terraform/external/pre-user-data -v $(pwd)/post-user-data:/usr/local/bin/provisioner/terraform/external/post-user-data --env-file pce.env artifacts.msap.io/mulesoft/core-paas-private-cloud-provisioner:v4.0.13 cluster-provision ## Open Port 61009 Before Installation If you are installing Anypoint Platform PCE using the GUI-based installer, you must enable this port before running the installer. You must open this port in the cluster’s security group before running the installer using AWS Web Console. ## Destroy Resources You can destroy all resources that were created using the provisioner by running the `cluster-deprovision` command, as shown in the following example. Ensure the `TELEKUBE_CLUSTER_NAME` environment variable in your environment file has the correct value for the target cluster you want to destroy: docker run --rm --env-file pce.env artifacts.msap.io/mulesoft/core-paas-private-cloud-provisioner:v4.0.13 cluster-deprovision ## Troubleshooting Review the following information in the event you experience problems when creating AWS resources. - 403 Forbidden Check that your keys policies are not denying access to any EC2 or S3 resource. You should be able to run the following basic command using AWS CLI with your keys: aws sts get-caller-identity - Limit exceeded The AWS account might have some limits set on the amount of resources you can create. Delete some unused resources or request a limit increase on AWS. - AMI not found Make sure you are using the AMI name as the value for the environment variable and not the ID. ## See Also - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) - [Install Terraform](https://developer.hashicorp.com/terraform/tutorials/aws-get-started/install-cli) - [Linux Bastion Hosts on the AWS Cloud: Quick Start Reference Deployment](https://docs.aws.amazon.com/quickstart/latest/linux-bastion/welcome.html) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure a Disclaimer for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can add a custom message to your login page that users in your organization see every time they log in: 1. From Anypoint Platform, select **Access Management** > **Disclaimer**. 2. Enter your disclaimer text. 3. Select **Save**. ![on prem features d22af](/mulesoft/current/en-us/private-cloud/3.2/_images/on-prem-features-d22af.png) You can also set up a custom HTML message. --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure Alerts for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) provides built-in alerts that are triggered when a condition specified in any alert definition is detected. Measurements are stored in [Prometheus](https://prometheus.io/docs/introduction/overview/) and read by [Alertmanager](https://prometheus.io/docs/alerting/latest/alertmanager/). Alertmanager sends emails when an alert is triggered. ## Alert Definitions Default alerts are listed in the following table:
Table 1. Alert Definitions
ComponentAlertDescription

CPU

High CPU usage

Triggers a warning when > 75% used; triggers a critical error when > 90% used

Memory

High memory usage

Triggers a warning when > 80% used; triggers a critical error when > 90% used

Systemd

Overall systemd health

Triggers an error when systemd detects a failed service

Systemd

Individual systemd unit health

Triggers an error when a systemd unit is not loaded or active

Filesystem

High disk space usage

Triggers a warning when > 80% used; triggers a critical error when > 90% used

Filesystem

High inode usage

Triggers a warning when > 90% used; triggers a critical error when > 95% used

System

Uptime

Triggers a warning when the uptime for a node is less than five minutes

System

Kernel parameters

Triggers an error if a parameter is not set. See value matrix for details.

Etcd

Etcd instance health

Triggers an error when an etcd leader is down longer than five minutes

Etcd

Etcd latency check

Triggers a warning when follower <→ leader latency exceeds 500 ms; triggers an error when it exceeds one second over a period of one minute

Docker

Docker daemon health

Triggers an error when the Docker daemon is down

Kubernetes

Kubernetes node readiness

Triggers an error when the node is not ready

## Configure Alert Definitions You define new alerts using a gravity resource called `alert`, as shown in the following example: ```copy kind: alert version: v2 metadata: name: cpu-alert spec: # the alert name alert_name: CPUAlert # the rule group the alert belongs to group_name: test-group # the alert expression formula: | node:cluster_cpu_utilization:ratio * 100 > 80 # the alert labels labels: severity: info # the alert annotations annotations: description: | Cluster CPU usage exceeds 80%. ``` See the [Alerting Rules](https://prometheus.io/docs/prometheus/latest/configuration/alerting_rules/) documentation for more details about Prometheus alerts. - To create an alert, run: gravity resource create alert.yaml - To view existing alerts, run: gravity resource get alerts - To remove an alert, run: gravity resource rm alert cpu-alert ## Configure Alerts Delivery To configure Alertmanager to send email alerts, create the following gravity resources: 1. Using the following spec, create a file named `smtp-config.yaml` inside gravity, and replace the placeholder values with those of your SMTP configuration: ```copy kind: smtp version: v2 metadata: name: smtp spec: host: port: username: password: --- kind: alerttarget version: v2 metadata: name: email-alerts spec: # email address of the alerts recipient email: ``` 2. Run `gravity resource create smtp-config.yaml`. You should see the following output: Created cluster SMTP configuration Created monitoring alert target "email-alerts" 3. Add a default router to the Alertmanager configuration inside gravity: kubectl get secret -n monitoring alertmanager-monitoring-kube-prometheus-alertmanager -o json | jq --arg foo "$(kubectl get secret -n monitoring alertmanager-monitoring-kube-prometheus-alertmanager -o json | jq -r '.data\["alertmanager.yaml"\]' | base64 -d | yq r - --tojson | jq -r '.route.routes\[1\] |= . + {"match":{"alertname": "Watchdog", "receiver": "default", "continue": true}}' | jq -r '.route.routes\[0\].match += {"continue":true}' | yq r - -P | base64 | tr -d '\\n')" '.data\["alertmanager.yaml"\]=$foo' | kubectl apply -f - 4. Configure the FROM email address by replacing the value: kubectl get secret -n monitoring alertmanager-monitoring-kube-prometheus-alertmanager -o json | jq --arg foo "$(kubectl get secret -n monitoring alertmanager-monitoring-kube-prometheus-alertmanager -o json | jq -r '.data\["alertmanager.yaml"\]' | base64 -d | yq w - 'global.smtp\_from' | base64 | tr -d '\\n')" '.data\["alertmanager.yaml"\]=$foo' | kubectl apply -f - 5. Restart Alertmanager pods: kubectl delete pod -n monitoring -l app=alertmanager 6. Test Alertmanager by running the following command inside gravity: curl -H 'Content-Type: application/json' -d '\[{"labels":{"alertname":"test-alert","state":"firing"}}\]' http://monitoring-kube-prometheus-alertmanager.monitoring.svc.cluster.local:9093/api/v1/alerts ## Troubleshooting Alerts Common troubleshooting tasks include the following: - Verify that your SMTP server can send and receive emails using the addresses you defined as the `FROM` and `TO` addresses when you configured alerts delivery. - Verify that your cluster nodes can communicate with your SMTP server. For example, use `telnet` to connect to your SMTP server from one of your cluster nodes: telnet my.smtp.server.com 587 Trying XXX.XXX.XXX.XXX... Connected to my.smtp.server.com. Escape character is '^\]'. 220 my.smtp.server.com ESMTP ^\[^\] telnet> quit Connection closed. ## See Also - [Ops Center](https://docs.mulesoft.com/private-cloud/3.2/managing-via-the-ops-center) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure an External Identity Provider in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires an external identity provider. All users must be registered via external identity except the admin you register in the installation wizard. The following external identity providers are supported: - [Configure LDAP in Anypoint Platform Private Cloud](https://docs.mulesoft.com/access-management/conf-ldap-private-cloud-task) - [Configure Ping Federate](https://docs.mulesoft.com/access-management/conf-client-mgmt-pf-task) - [Configure SAML](https://docs.mulesoft.com/access-management/conf-saml-sso) - [Configure OpenID Connect](https://docs.mulesoft.com/access-management/conf-openid-connect-task) For troubleshooting information, see [Anypoint Platform Single Sign-On (SSO) using OpenID Connect (OIDC) Troubleshooting Guide](https://help.salesforce.com/s/articleView?id=001114931&type=1). --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After you complete Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation, review the following tasks, many of which are required before you can successfully deploy and manage applications, register APIs, and use API Exchange on an Anypoint Platform PCE environment running on your own local network. - Optional: [Delete the Local User after Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/install-disable-local-user) - [Configure an External Identity Provider in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/install-config-ldap-pce) - [Configure SMTP in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/access-management-SMTP) - [Configure DNS in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/access-management-dns) - [Configure Security in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/access-management-security) - [Configure a Disclaimer in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/access-management-disclaimer) > [!NOTE] At the time of this writing, some Docker containers must run as root. Processes running as > root do not expose functionality outside the cluster. Processes that serve external traffic do not > run as root, to maximize security. ## See Also - [About Servers, Server Groups, and Clusters](https://docs.mulesoft.com/runtime-manager/managing-servers) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure Anypoint Studio for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To configure Anypoint Studio 7.15 for Anypoint Platform Private Cloud Edition 3.2, follow the steps in [Configure Anypoint PCE in Studio](https://docs.mulesoft.com/studio/pce-configuration). --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure Backup and Restore for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To preserve your data in case of system failure or problems with other operations, such as patching or performing upgrades, ensure that your system is automatically backed up. Regular backups are extremely important for disaster recovery procedures. Configure your backup to run at least once per day; or more often depending on your requirements. Store your backup archive on external storage outside the Anypoint Platform cluster. Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires you to install, configure, and manage an NFS server. Ensure that you configure and perform frequent backups of this server. > [!IMPORTANT] Backup and restore for Anypoint Monitoring and Anypoint Visualizer are perfored > outside the normal Gravity backup and restore mechanism, and require a separate procedure. ## Create a Backup To create a backup, run the following command on any node in the cluster: gravity backup /var/lib/gravity/planet/share/backup.tar.gz This command creates an archive of the current system state in: `/var/lib/gravity/planet/share/backup.tar.gz`. You can pass the following optional parameters to the gravity backup command: - `--follow` Outputs the backup job logs to stdout. In the following example, `tee` enables you to output to stdout and file at the same time: gravity backup /var/lib/gravity/planet/share/backup.tar.gz --follow | tee -a /var/lib/gravity/planet/share/backup.log - `--timeout` Specifies the deadline for the backup job, for example, 30s or 5m. Default is 20m. Backup contents:
Table 1. Backup Contents
InformationIncluded in PCE Backup

Databases

Y

Configurations

Y

Secrets

Y

Object Store

Y

Disclaimer settings

Y

SMTP settings

Y

DNS settings

N

Platform certificate

N

Ops Center certificate

N

NFS settings

N

NFS files

N

Drives

N

Runtimes*

N

Anypoint Monitoring metrics

N

\*Anypoint Platform PCE has information only about runtimes, such as their names and status. ## Perform a System Restore To restore a system from your backup archive, use the original cluster or create a new cluster. > [!WARNING] Some settings, including NFS, certificates, and DNS, are not backed up and must be > configured for the environment in which the restore will be performed before running the restore. > [!WARNING] You cannot perform a backup between different versions of Anypoint Platform PCE. The > new installation you create must use the same version of Anypoint Platform PCE as your backup. 1. Log in to one of the nodes in your cluster. 2. Move the compressed backup archive file to a folder on any main node in the environment to be restored. You can transfer this file securely by using the following command: scp /backup-path/to-restore.tar.gz your\_username@remotehost.edu:/var/lib/gravity/planet/share 3. Restore the cluster from the backup archive: sudo gravity restore /var/lib/gravity/planet/share/to-restore.tar.gz You can pass the following optional parameters: - `--follow` Outputs the restore job logs to stdout. - `--timeout` Specifies the maximum time allowed for the restore operation, for example, 30s, 5m, and so on. The default value is `80m`. If you have a very large number of applications and files, use a minimum of `160m` as the `--timeout` value. 4. Wait for the operation to complete, which is typically 40 to 60 minutes. 5. Manually restore the NFS files if you are not reusing the original NFS. ## Anypoint Monitoring and Anypoint Visualizer If Anypoint Monitoring and Anypoint Visualizer are enabled, follow these procedures to back up and restore them. Anypoint Monitoring and Anypoint Visualizer backup and restore are outside the normal Gravity backup and restore mechanism, and require additional procedures. These components store metrics and other information, but they do not store platform configuration. Platform configuration information is handled by the system backup and restore procedures. Anypoint Monitoring and Anypoint Visualizer must back up the data of the component InfluxDB (`dias-prov-k8s-am-influxdb-comp`). Anypoint Monitoring and Anypoint Visualizer both store their configuration data in a Postgres database. The default Gravity mechanism backs up and restores this database. Anypoint Monitoring and Anypoint Visualizer requires 4 TB volumes for `amv` nodes. A full InfluxDB backup can be up to 8 TB of data. (Only two data nodes are operative.) You must provide enough free space on the target directory for the entire backup. Restore time is proportional to the size of the backup. During restore, existing data on the target cluster will be backed up and erased upon user approval. The backup and restore script must be run on a main node and must have `sudo` access because it uses `kubectl`. The script validates the restore disk size, databases, measurements, retention policies, and series cardinality for all measurements in each database. ### Procedure The InfluxDB backup and restore script is in the Anypoint Platform PCE environment: `/var/lib/gravity/site/packages/unpacked/gravitational.io/anypoint/3.2.0*/resources/kubernetes/amv-backup-restore/amv-backup-restore.sh` The script has these parameters: -a : (Required) action to perform : backup or restore. \[string\]. Default NO DEFAULT ACTION. -va : (optional) verbose output. Default non-verbose. -f : (Required) name of the tar file. Default for backup : dias-prov-k8s-am-influxdb-comp\_YYYYMMD. \[string\]. Default for restore : /var/lib/data/influxdb/backup/dias-prov-k8s-am-influxdb-comp\_YYYYMMDD.tar.gz. \[Full Path\]. -d : (Required for backup) full path to location where backup tar file is stored. Default for backup : /var/lib/data/influxdb/backup. \[Full Path\]. Default for restore : same as tar file path. \[Full Path\]. -s : (optional) Optional sub-component filter, use only for backup of only meta or only data. -h : Print HELP. ### Examples Verbose backup: `sudo bash amv-backup-restore.sh -va backup -f 2020-01-01-influxdb-backup -d path/to/backup/dir` Non-verbose restore: `sudo bash amv-backup-restore.sh -a restore -f /var/lib/data/influxdb/backup/dias-prov-k8s-am-influxdb-comp_YYYYMMDD.tar.gz` ## See Also - [Install Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/3.2/install-workflow) - [Upgrade Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/3.2/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure DNS on Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you did not perform this step before installation, you must provide the DNS or IP address where your Anypoint Platform Private Cloud Edition installation is exposed to the local network. This is required for Mule to communicate with Anypoint Platform. 1. From Anypoint Platform, click Access Management. 2. Click the **DNS/IP** tab. 3. In the **Platform DNS/IP** field, enter the address where you are hosting the platform. 4. Click **Save**. > [!NOTE] \== If you perform a DNS update that changes the endpoints that Mule instances use to > connect to Runtime Manager, you must restart the Mule instances before they can connect to the new > endpoints. == --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure Security for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To use Anypoint Platform, you must provide TLS credentials. You can upload a certificate through the Anypoint Platform UI using access management, and configure Anypoint Platform to use this certificate to ensure secure communication within the platform. This certificate must be trusted by every machine connected to the platform, including Mule runtime engines. You can also provide a custom cipher and DH groups for the key exchange method, if required. When you replace or change the certificate, Anypoint Platform restarts the nginx server. ![on prem features 5f98d](/mulesoft/current/en-us/private-cloud/3.2/_images/on-prem-features-5f98d.png) 1. Create a certificate. You can use a tool like OpenSSL to create your certificate. To work with Anypoint Platform, your certificate must be in an `nginx` compatible format. The certificate must contain the domain name of your installation as the CN (Common Name) and as the SAN (Subject Alternative Name). 2. From Anypoint Platform, select Access Management. 3. In the left navigation bar, select Security. 4. In the Certificate field, select Browse to select the certificate on your local system. 5. In the Key field, select Browse to select the private key for your certificate. 6. For additional security, you can upload your own cipher and Diffie-Hellman group file. Note that this step is optional. 1. Deselect **Use default ciphers and DH Group**. 2. Enter a set of ciphers from the list of valid NGINX ciphers. If you are using Open SSL, you can view this list by running the `openssl ciphers` command. The cipher you enter must be in an `nginx` compatible format as shown in the following example: ECDHE-RSA-AES256-GCM-SHA512:DHE-RSA-AES256-GCM-SHA512:ECDHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-SHA384 3. Generate `dhparam.pem` (the Diffie-Hellman group file) using the following `openssl` command: openssl dhparam 2048 -out dhparam.pem 4. Upload the `dhparam.pem` file from your system. --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Configure SMTP in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Because Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) does not provide an SMTP server, you must set up your own SMTP email server to deliver email alerts generated by API Manager and Runtime Manager. The SMTP email server also sends email messages for platform infrastructure alerts, including CPU, memory, and disk alerts. You must direct your local installation of Anypoint Platform PCE to this server from the Access Management panel. > [!NOTE] Only SMTP servers with Authentication are supported (anonymous SMTP relay servers are not > supported). > SMTPS is not supported. ## Connect to an Existing SMTP Server 1. From Anypoint Platform, select **Access Management**. 2. Select **SMTP**. 3. Provide the following information for your SMTP server: ![on premises features 2c60a](/mulesoft/current/en-us/private-cloud/3.2/_images/on-premises-features-2c60a.png)
Table 1. SMTP Server Parameters
FieldDescriptionExample

Host

The host name of your SMTP server

smtp.gmail.com

Port

The port used to communicate to your SMTP server

587

Username

The username to authenticate to your SMTP server

mulesoft.onprem.smtp@gmail.com

Password

Your SMTP password

examplepassphrase

> [!NOTE] To use anonymous logins for your SMTP server, leave the **Username** and **Password** > fields empty. 4. After entering your configuration, select **Save**. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring a Disclaimer for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To configure a disclaimer for your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), add a custom message to your login page that users in your organization see every time they log in: 1. From Anypoint Platform, select **Access Management** > **Disclaimer**. 2. Enter your disclaimer text (you can also set up a custom HTML message). 3. Select **Save**. ![Disclaimer window to enter text under the Access Mananagemet UI](/mulesoft/current/en-us/private-cloud/4.0/_images/on-prem-features-d22af.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring a Disclaimer for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To configure a disclaimer for your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), add a custom message to your login page that users in your organization see every time they log in: 1. From Anypoint Platform, select **Access Management** > **Disclaimer**. 2. Enter your disclaimer text (you can also set up a custom HTML message). 3. Select **Save**. ![Disclaimer window to enter text under the Access Mananagemet UI](/mulesoft/current/en-us/private-cloud/4.1/_images/on-prem-features-d22af.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring a Disclaimer for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To configure a disclaimer for your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), add a custom message to your login page that users in your organization see every time they log in: 1. From Anypoint Platform, select **Access Management** > **Disclaimer**. 2. Enter your disclaimer text (you can also set up a custom HTML message). 3. Select **Save**. ![Disclaimer window to enter text under the Access Mananagemet UI](/mulesoft/current/en-us/private-cloud/4.2/_images/on-prem-features-d22af.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring a Disclaimer for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To configure a disclaimer for your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), add a custom message to your login page that users in your organization see every time they log in: 1. From Anypoint Platform, select **Access Management** > **Disclaimer**. 2. Enter your disclaimer text (you can also set up a custom HTML message). 3. Select **Save**. ![Disclaimer window to enter text under the Access Mananagemet UI](/mulesoft/current/en-us/private-cloud/latest/_images/on-prem-features-d22af.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring an External Identity Provider in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires an external identity provider. All users must be registered via external identity except the admin you register in the installation wizard. The following external identity providers are supported: - [Configuring LDAP in Anypoint Platform PCE](https://docs.mulesoft.com/access-management/conf-ldap-private-cloud-task) - [Configuring Ping Federate](https://docs.mulesoft.com/access-management/conf-client-mgmt-pf-task) - [Configuring SAML](https://docs.mulesoft.com/access-management/conf-saml-sso) - [Configuring OpenID Connect](https://docs.mulesoft.com/access-management/conf-openid-connect-task) For troubleshooting information, see [Anypoint Platform Single Sign-On (SSO) using OpenID Connect (OIDC) Troubleshooting Guide](https://help.salesforce.com/s/articleView?id=001114931&type=1). ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring an External Identity Provider in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires an external identity provider. All users must be registered via external identity except the admin you register in the installation wizard. The following external identity providers are supported: - [Configuring LDAP in Anypoint Platform PCE](https://docs.mulesoft.com/access-management/conf-ldap-private-cloud-task) - [Configuring Ping Federate](https://docs.mulesoft.com/access-management/conf-client-mgmt-pf-task) - [Configuring SAML](https://docs.mulesoft.com/access-management/conf-saml-sso) - [Configuring OpenID Connect](https://docs.mulesoft.com/access-management/conf-openid-connect-task) For troubleshooting information, see [Anypoint Platform Single Sign-On (SSO) using OpenID Connect (OIDC) Troubleshooting Guide](https://help.salesforce.com/s/articleView?id=001114931&type=1). ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring an External Identity Provider in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires an external identity provider. All users must be registered via external identity except the admin you register in the installation wizard. The following external identity providers are supported: - [Configuring LDAP in Anypoint Platform PCE](https://docs.mulesoft.com/access-management/conf-ldap-private-cloud-task) - [Configuring Ping Federate](https://docs.mulesoft.com/access-management/conf-client-mgmt-pf-task) - [Configuring SAML](https://docs.mulesoft.com/access-management/conf-saml-sso) - [Configuring OpenID Connect](https://docs.mulesoft.com/access-management/conf-openid-connect-task) For troubleshooting information, see [Anypoint Platform Single Sign-On (SSO) using OpenID Connect (OIDC) Troubleshooting Guide](https://help.salesforce.com/s/articleView?id=001114931&type=1). ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring an External Identity Provider in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires an external identity provider. All users must be registered via external identity except the admin you register in the installation wizard. The following external identity providers are supported: - [Configuring LDAP in Anypoint Platform PCE](https://docs.mulesoft.com/access-management/conf-ldap-private-cloud-task) - [Configuring Ping Federate](https://docs.mulesoft.com/access-management/conf-client-mgmt-pf-task) - [Configuring SAML](https://docs.mulesoft.com/access-management/conf-saml-sso) - [Configuring OpenID Connect](https://docs.mulesoft.com/access-management/conf-openid-connect-task) For troubleshooting information, see [Anypoint Platform Single Sign-On (SSO) using OpenID Connect (OIDC) Troubleshooting Guide](https://help.salesforce.com/s/articleView?id=001114931&type=1). ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring Anypoint Code Builder for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Use [Anypoint Code Builder](https://docs.mulesoft.com/anypoint-code-builder/start-acb-desktop) in VS Code to design, develop, and deploy applications to your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) environment. You can use Anypoint Code Builder without MuleSoft Vibes by following the standard Anypoint Code Builder documentation. MuleSoft Vibes is optional. ## Before You Begin Before you start developing with Anypoint Code Builder, ensure that you have: - An Anypoint Platform PCE 4.3 environment - Access to an [Anypoint Platform](https://anypoint.mulesoft.com) account on anypoint.mulesoft.com This access is separate from your Anypoint Platform PCE license. Anypoint Code Builder uses it to retrieve assets and export the application artifact. - Mule runtime engine 4.9 LTS or 4.12 Edge support. - Microsoft Visual Studio Code (VS Code) installed on your local machine - Anypoint CLI installed and configured for your Anypoint Platform PCE instance For installation instructions, see [Installing Anypoint CLI](https://docs.mulesoft.com/anypoint-cli/latest/install). - These Anypoint CLI plugins installed by running the following command once for each plugin: anypoint-cli-v4 plugins:install Replace `` with each of the following: - [anypoint-cli-account-plugin](https://docs.mulesoft.com/anypoint-cli/latest/account) - [anypoint-cli-api-manager-plugin](https://docs.mulesoft.com/anypoint-cli/latest/api-mgr) - [anypoint-cli-exchange-plugin](https://docs.mulesoft.com/anypoint-cli/latest/exchange-assets) - [mulesoft-anypoint-cli-designcenter-plugin](https://docs.mulesoft.com/anypoint-cli/latest/design-center) MuleSoft Vibes is optional. If you use it, install [MuleSoft Vibes Skills](https://docs.mulesoft.com/anypoint-code-builder/vibes-skills) on your local machine, in Claude, Copilot, or another supported developer agent. > [!NOTE] Contact your MuleSoft representative for details on how to configure MuleSoft Vibes within > your environment. > > For installation instructions, see > [Using Skills with MuleSoft Vibes](https://docs.mulesoft.com/anypoint-code-builder/vibes-skills#install-mulesoft-skills-in-claude-code). ## Connect Anypoint Platform PCE with Anypoint Code Builder To configure Anypoint Platform PCE with Anypoint Code Builder, connect your local VS Code environment to your Anypoint Platform account: 1. In VS Code, navigate to **Settings** > **Extensions** > **Anypoint Code Builder (Platform)**. 2. Set the **Mule Control Plane** dropdown to **Private Cloud Edition**, and enter your PCE URL under **Mule: Control Plane PCE Url**. For example: `[https://pce-instance.elb.us-east-2.amazonaws.com/](https://pce-instance.elb.us-east-2.amazonaws.com/)` 3. Optional: Install and configure [MuleSoft Vibes Skills](https://docs.mulesoft.com/anypoint-code-builder/vibes-skills#install-mulesoft-skills-in-claude-code). 4. Open the **Command Palette**, run **MuleSoft: Login to Anypoint Platform**, and complete authentication in your browser. 5. If you use SSO or MFA, set the `ANYPOINT_BEARER_TOKEN` environment variable in your terminal to an access token from your authenticated Anypoint Platform session. After you complete this setup, you can [deploy Mule applications](https://docs.mulesoft.com/private-cloud/latest/deploy-apps-code-builder) to Anypoint Platform PCE. ## See Also - [Deploying Mule Applications to Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/deploy-apps) - [Deploying Mule Applications Using Anypoint Code Builder](https://docs.mulesoft.com/private-cloud/latest/deploy-apps-code-builder) - [Developing Integrations](https://docs.mulesoft.com/anypoint-code-builder/int-developing-integrations) - [Designing API Specs and Fragments](https://docs.mulesoft.com/anypoint-code-builder/des-designing-api-specs) - [Configuring Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-studio) - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After you complete the Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation, review the following task to deploy and manage applications, register APIs, and use API Exchange on an Anypoint Platform PCE environment running on your own local network. - [Deleting the Local User after Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/install-disable-local-user) - [Configuring an External Identity Provider in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/install-config-ldap-pce) - [Configuring SMTP in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/access-management-SMTP) - [Configuring DNS in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/access-management-dns) - [Configuring Security in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/access-management-security) - [Configuring a Disclaimer in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/access-management-disclaimer) ## See Also - [About Servers, Server Groups, and Clusters](https://docs.mulesoft.com/runtime-manager/managing-servers) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After you complete the Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation, review the following task to deploy and manage applications, register APIs, and use API Exchange on an Anypoint Platform PCE environment running on your own local network. - [Deleting the Local User after Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/install-disable-local-user) - [Configuring an External Identity Provider in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/install-config-ldap-pce) - [Configuring SMTP in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/access-management-SMTP) - [Configuring DNS in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/access-management-dns) - [Configuring Security in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/access-management-security) - [Configuring a Disclaimer in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/access-management-disclaimer) - [Limiting Concurrent Login in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/access-management-concurrent-login) ## See Also - [About Servers, Server Groups, and Clusters](https://docs.mulesoft.com/runtime-manager/managing-servers) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After you complete the Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation, review the following task to deploy and manage applications, register APIs, and use API Exchange on an Anypoint Platform PCE environment running on your own local network. - [Deleting the Local User after Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/install-disable-local-user) - [Configuring an External Identity Provider in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/install-config-ldap-pce) - [Configuring SMTP in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/access-management-SMTP) - [Configuring DNS in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/access-management-dns) - [Configuring Security in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/access-management-security) - [Configuring a Disclaimer in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/access-management-disclaimer) - [Limiting Concurrent Login in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/access-management-concurrent-login) ## See Also - [About Servers, Server Groups, and Clusters](https://docs.mulesoft.com/runtime-manager/managing-servers) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After you complete the Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation, review the following task to deploy and manage applications, register APIs, and use API Exchange on an Anypoint Platform PCE environment running on your own local network. - [Deleting the Local User after Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/install-disable-local-user) - [Configuring an External Identity Provider in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/install-config-ldap-pce) - [Configuring SMTP in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/access-management-SMTP) - [Configuring DNS in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/access-management-dns) - [Configuring Security in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/access-management-security) - [Configuring a Disclaimer in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/access-management-disclaimer) - [Limiting Concurrent Login in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/access-management-concurrent-login) ## See Also - [About Servers, Server Groups, and Clusters](https://docs.mulesoft.com/runtime-manager/managing-servers) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring Anypoint Studio for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Studio 7.19 and later adds support for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0.0. ## Configure Your PCE Certificates into Studio Integrate your Studio installation with the Anypoint Platform components in an Anypoint Platform PCE deployment. ### Determine the Java Version and Location 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Java** > **Installed JRES**. 3. Check the location of the JRE you use: Java 8 or Java 17. ![JRE location in Anypoint Studio](/mulesoft/current/en-us/studio/_images/pce4-jre-location.png) ### Export the Certificate Using the Platform DNS Run the following command by replacing `` with your actual server URL: `keytool -printcert -sslserver :443 -rfc > /tmp/PCE_Certificate.cer` For example: ```xml keytool -printcert -sslserver $ANYPOINT_DNS:443 -rfc > /tmp/PCE_Certificate.cer ``` Starting with Anypoint Studio 7.17, the default JDK version for Studio is Java 17. To ensure proper functionality, add the certificate to the default JDK version that Studio supports. - If you use an external JDK, add the certificate to the external JDK that you use with Studio. - If you use the default JDK Studio provides, add the certificate to the default JDK, regardless of the JDK version you use in your project. After installing the certificate, if you have Studio open, restart it. ### Import the Certificate Run the following command by replacing `` with the path to your embedded Java in Studio: `sudo keytool -import -trustcacerts -keystore /lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` For example, if you’re using the embedded Java 17 on macOS: ```xml sudo keytool -import -trustcacerts -keystore /Applications/AnypointStudio.app/Contents/Eclipse/plugins/org.mule.tooling.jdk.macosx.x86_64_1.3.0/Contents/Home/jre/lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` ``` For example, if you’re using the embedded Java 17 on Windows: ```xml keytool -import -trustcacerts -keystore C:[...]\AnypointStudio\plugins\org.mule.tooling.jdk.win32.x86_64_1.3.3\lib\security\cacerts -storepass changeit -file /tmp/PCE_Certificate.cer (1) ```
1The path changes with each Studio version because the JDK plugin version is included in the Windows path, and Studio includes a different JDK plugin version with each release.
### Set a Private Cloud Instance URL 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Anypoint Studio** > **Anypoint Platform**. ![Anypoint Studio highlighted in the navigation bar.](/mulesoft/current/en-us/reuse/_images/studio-anypoint-platform-settings.png) 3. Select the **Use a Private Cloud Instance URL**. 4. Type the URL of your PCE installation in **Private Cloud Instance URL**. ![\*Private Cloud Instance URL\* field highlighted in the in the \*PCE Settings\* section.](/mulesoft/current/en-us/studio/_images/pce-properties.png) 5. Click **Apply and Close**. 6. Restart Studio. ## Configure your PCE User Login The first time you configure your user’s login, Studio asks you to confirm your PCE certificate: 1. Open the preferences view. Either navigate to **Anypoint Studio** > **Preferences…​** or press `cmd + ,`. 2. Expand **Anypoint Studio** in the left navigation bar, select **Authentication** > **Add**. ![\*Authentication\* option and \*Add\* button highlighted in the \*Preference\* window.](/mulesoft/current/en-us/studio/_images/preferences-authentication-add.png) 3. Select **Yes** in the **Invalid Certificate** message that appears. ![\*Yes\* button highlighted in the \*Invalid Certificate\* message window.](/mulesoft/current/en-us/studio/_images/invalid-certificate-pce.png) 4. Add your PCE credentials and sign-in. 5. Select **Apply and Close**. 6. Restart Studio. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring Anypoint Studio for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Studio 7.19 and later adds support for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0.0. ## Configure Your PCE Certificates into Studio Integrate your Studio installation with the Anypoint Platform components in an Anypoint Platform PCE deployment. ### Determine the Java Version and Location 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Java** > **Installed JRES**. 3. Check the location of the JRE you use: Java 8 or Java 17. ![JRE location in Anypoint Studio](/mulesoft/current/en-us/studio/_images/pce4-jre-location.png) ### Export the Certificate Using the Platform DNS Run the following command by replacing `` with your actual server URL: `keytool -printcert -sslserver :443 -rfc > /tmp/PCE_Certificate.cer` For example: ```xml keytool -printcert -sslserver $ANYPOINT_DNS:443 -rfc > /tmp/PCE_Certificate.cer ``` Starting with Anypoint Studio 7.17, the default JDK version for Studio is Java 17. To ensure proper functionality, add the certificate to the default JDK version that Studio supports. - If you use an external JDK, add the certificate to the external JDK that you use with Studio. - If you use the default JDK Studio provides, add the certificate to the default JDK, regardless of the JDK version you use in your project. After installing the certificate, if you have Studio open, restart it. ### Import the Certificate Run the following command by replacing `` with the path to your embedded Java in Studio: `sudo keytool -import -trustcacerts -keystore /lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` For example, if you’re using the embedded Java 17 on macOS: ```xml sudo keytool -import -trustcacerts -keystore /Applications/AnypointStudio.app/Contents/Eclipse/plugins/org.mule.tooling.jdk.macosx.x86_64_1.3.0/Contents/Home/jre/lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` ``` For example, if you’re using the embedded Java 17 on Windows: ```xml keytool -import -trustcacerts -keystore C:[...]\AnypointStudio\plugins\org.mule.tooling.jdk.win32.x86_64_1.3.3\lib\security\cacerts -storepass changeit -file /tmp/PCE_Certificate.cer (1) ```
1The path changes with each Studio version because the JDK plugin version is included in the Windows path, and Studio includes a different JDK plugin version with each release.
### Set a Private Cloud Instance URL 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Anypoint Studio** > **Anypoint Platform**. ![Anypoint Studio highlighted in the navigation bar.](/mulesoft/current/en-us/reuse/_images/studio-anypoint-platform-settings.png) 3. Select the **Use a Private Cloud Instance URL**. 4. Type the URL of your PCE installation in **Private Cloud Instance URL**. ![\*Private Cloud Instance URL\* field highlighted in the in the \*PCE Settings\* section.](/mulesoft/current/en-us/studio/_images/pce-properties.png) 5. Click **Apply and Close**. 6. Restart Studio. ## Configure your PCE User Login The first time you configure your user’s login, Studio asks you to confirm your PCE certificate: 1. Open the preferences view. Either navigate to **Anypoint Studio** > **Preferences…​** or press `cmd + ,`. 2. Expand **Anypoint Studio** in the left navigation bar, select **Authentication** > **Add**. ![\*Authentication\* option and \*Add\* button highlighted in the \*Preference\* window.](/mulesoft/current/en-us/studio/_images/preferences-authentication-add.png) 3. Select **Yes** in the **Invalid Certificate** message that appears. ![\*Yes\* button highlighted in the \*Invalid Certificate\* message window.](/mulesoft/current/en-us/studio/_images/invalid-certificate-pce.png) 4. Add your PCE credentials and sign-in. 5. Select **Apply and Close**. 6. Restart Studio. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring Anypoint Studio for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Studio 7.19 and later adds support for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0.0. ## Configure Your PCE Certificates into Studio Integrate your Studio installation with the Anypoint Platform components in an Anypoint Platform PCE deployment. ### Determine the Java Version and Location 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Java** > **Installed JRES**. 3. Check the location of the JRE you use: Java 8 or Java 17. ![JRE location in Anypoint Studio](/mulesoft/current/en-us/studio/_images/pce4-jre-location.png) ### Export the Certificate Using the Platform DNS Run the following command by replacing `` with your actual server URL: `keytool -printcert -sslserver :443 -rfc > /tmp/PCE_Certificate.cer` For example: ```xml keytool -printcert -sslserver $ANYPOINT_DNS:443 -rfc > /tmp/PCE_Certificate.cer ``` Starting with Anypoint Studio 7.17, the default JDK version for Studio is Java 17. To ensure proper functionality, add the certificate to the default JDK version that Studio supports. - If you use an external JDK, add the certificate to the external JDK that you use with Studio. - If you use the default JDK Studio provides, add the certificate to the default JDK, regardless of the JDK version you use in your project. After installing the certificate, if you have Studio open, restart it. ### Import the Certificate Run the following command by replacing `` with the path to your embedded Java in Studio: `sudo keytool -import -trustcacerts -keystore /lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` For example, if you’re using the embedded Java 17 on macOS: ```xml sudo keytool -import -trustcacerts -keystore /Applications/AnypointStudio.app/Contents/Eclipse/plugins/org.mule.tooling.jdk.macosx.x86_64_1.3.0/Contents/Home/jre/lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` ``` For example, if you’re using the embedded Java 17 on Windows: ```xml keytool -import -trustcacerts -keystore C:[...]\AnypointStudio\plugins\org.mule.tooling.jdk.win32.x86_64_1.3.3\lib\security\cacerts -storepass changeit -file /tmp/PCE_Certificate.cer (1) ```
1The path changes with each Studio version because the JDK plugin version is included in the Windows path, and Studio includes a different JDK plugin version with each release.
### Set a Private Cloud Instance URL 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Anypoint Studio** > **Anypoint Platform**. ![Anypoint Studio highlighted in the navigation bar.](/mulesoft/current/en-us/reuse/_images/studio-anypoint-platform-settings.png) 3. Select the **Use a Private Cloud Instance URL**. 4. Type the URL of your PCE installation in **Private Cloud Instance URL**. ![\*Private Cloud Instance URL\* field highlighted in the in the \*PCE Settings\* section.](/mulesoft/current/en-us/studio/_images/pce-properties.png) 5. Click **Apply and Close**. 6. Restart Studio. ## Configure your PCE User Login The first time you configure your user’s login, Studio asks you to confirm your PCE certificate: 1. Open the preferences view. Either navigate to **Anypoint Studio** > **Preferences…​** or press `cmd + ,`. 2. Expand **Anypoint Studio** in the left navigation bar, select **Authentication** > **Add**. ![\*Authentication\* option and \*Add\* button highlighted in the \*Preference\* window.](/mulesoft/current/en-us/studio/_images/preferences-authentication-add.png) 3. Select **Yes** in the **Invalid Certificate** message that appears. ![\*Yes\* button highlighted in the \*Invalid Certificate\* message window.](/mulesoft/current/en-us/studio/_images/invalid-certificate-pce.png) 4. Add your PCE credentials and sign-in. 5. Select **Apply and Close**. 6. Restart Studio. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring Anypoint Studio for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Studio 7.19 and later adds support for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) version 4.0.0. ## Configure Your PCE Certificates into Studio Integrate your Studio installation with the Anypoint Platform components in an Anypoint Platform PCE deployment. ### Determine the Java Version and Location 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Java** > **Installed JRES**. 3. Check the location of the JRE you use: Java 8 or Java 17. ![JRE location in Anypoint Studio](/mulesoft/current/en-us/studio/_images/pce4-jre-location.png) ### Export the Certificate Using the Platform DNS Run the following command by replacing `` with your actual server URL: `keytool -printcert -sslserver :443 -rfc > /tmp/PCE_Certificate.cer` For example: ```xml keytool -printcert -sslserver $ANYPOINT_DNS:443 -rfc > /tmp/PCE_Certificate.cer ``` Starting with Anypoint Studio 7.17, the default JDK version for Studio is Java 17. To ensure proper functionality, add the certificate to the default JDK version that Studio supports. - If you use an external JDK, add the certificate to the external JDK that you use with Studio. - If you use the default JDK Studio provides, add the certificate to the default JDK, regardless of the JDK version you use in your project. After installing the certificate, if you have Studio open, restart it. ### Import the Certificate Run the following command by replacing `` with the path to your embedded Java in Studio: `sudo keytool -import -trustcacerts -keystore /lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` For example, if you’re using the embedded Java 17 on macOS: ```xml sudo keytool -import -trustcacerts -keystore /Applications/AnypointStudio.app/Contents/Eclipse/plugins/org.mule.tooling.jdk.macosx.x86_64_1.3.0/Contents/Home/jre/lib/security/cacerts -storepass changeit -file /tmp/PCE_Certificate.cer` ``` For example, if you’re using the embedded Java 17 on Windows: ```xml keytool -import -trustcacerts -keystore C:[...]\AnypointStudio\plugins\org.mule.tooling.jdk.win32.x86_64_1.3.3\lib\security\cacerts -storepass changeit -file /tmp/PCE_Certificate.cer (1) ```
1The path changes with each Studio version because the JDK plugin version is included in the Windows path, and Studio includes a different JDK plugin version with each release.
### Set a Private Cloud Instance URL 1. In Studio, go to **Anypoint Studio** > **Settings**. 2. Select **Anypoint Studio** > **Anypoint Platform**. ![Anypoint Studio highlighted in the navigation bar.](/mulesoft/current/en-us/reuse/_images/studio-anypoint-platform-settings.png) 3. Select the **Use a Private Cloud Instance URL**. 4. Type the URL of your PCE installation in **Private Cloud Instance URL**. ![\*Private Cloud Instance URL\* field highlighted in the in the \*PCE Settings\* section.](/mulesoft/current/en-us/studio/_images/pce-properties.png) 5. Click **Apply and Close**. 6. Restart Studio. ## Configure your PCE User Login The first time you configure your user’s login, Studio asks you to confirm your PCE certificate: 1. Open the preferences view. Either navigate to **Anypoint Studio** > **Preferences…​** or press `cmd + ,`. 2. Expand **Anypoint Studio** in the left navigation bar, select **Authentication** > **Add**. ![\*Authentication\* option and \*Add\* button highlighted in the \*Preference\* window.](/mulesoft/current/en-us/studio/_images/preferences-authentication-add.png) 3. Select **Yes** in the **Invalid Certificate** message that appears. ![\*Yes\* button highlighted in the \*Invalid Certificate\* message window.](/mulesoft/current/en-us/studio/_images/invalid-certificate-pce.png) 4. Add your PCE credentials and sign-in. 5. Select **Apply and Close**. 6. Restart Studio. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring Backup and Restore for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires you to install, configure, and manage an NFS server, ensuring frequent backups are configured and performed. To preserve your data: - Enable automatic backups to protect against system failures or operational issues. - Schedule backups at least daily, or more frequently based on your needs. - Store backups externally, outside the Anypoint Platform cluster. The backup and restore process is controlled via Anypoint Platform PCE APIs. ## Create a Backup To create a backup, run the following command: `curl -k /platform/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz", "nfs-server": "${fileSystemDns}", "nfs-path": "/"}'` ### Backup Content
InformationIncluded in Anypoint Platform PCE Backup

Databases

Yes

Configurations

Yes

Secrets

Yes

Object Store

Yes

Disclaimer settings

Yes

SMTP settings

Yes

DNS settings

No

Platform certificate

No

Ops Center certificate

No

NFS settings

No

NFS files

No

Drives

No

Runtimes*

No

Anypoint Monitoring metrics

No

\*Anypoint Platform PCE has information only about runtimes, such as their names and status. ## Perform a System Restore Consider the following before you perform a system restore: - Before running a restore, configure certain settings, such as NFS, certificates, and DNS because they are not included in the backup. - Backups cannot be transferred between different versions of Anypoint Platform PCE; therefore, your new installation must match the version of your backup. To perform the restore, run the following command: `curl -k /platform/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz", "nfs-server": "${fileSystemDns}", "nfs-path": "/"}` The backup file is stored on the NFS storage. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) - [Installing Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) - [Upgrade Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/4.0/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring Backup and Restore for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires you to install, configure, and manage an NFS server, ensuring frequent backups are configured and performed. To preserve your data: - Enable automatic backups to protect against system failures or operational issues. - Schedule backups at least daily, or more frequently based on your needs. - Store backups externally, outside the Anypoint Platform cluster. The backup and restore process is controlled via Anypoint Platform PCE APIs. ## Create a Backup To create a backup, run the following command: `curl -k /platform/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz", "nfs-server": "${fileSystemDns}", "nfs-path": "/"}'` ### Backup Content
InformationIncluded in Anypoint Platform PCE Backup

Databases

Yes

Configurations

Yes

Secrets

Yes

Object Store

Yes

Disclaimer settings

Yes

SMTP settings

Yes

DNS settings

No

Platform certificate

No

NFS settings

No

NFS files

No

Drives

No

Runtimes*

No

Anypoint Monitoring metrics

No

\*Anypoint Platform PCE has information only about runtimes, such as their names and status. ## Perform a System Restore Consider the following before you perform a system restore: - Before running a restore, configure certain settings, such as NFS, certificates, and DNS because they are not included in the backup. - Backups cannot be transferred between different versions of Anypoint Platform PCE; therefore, your new installation must match the version of your backup. To perform the restore, run the following command: `curl -k /platform/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz", "nfs-server": "${fileSystemDns}", "nfs-path": "/"}` The backup file is stored on the NFS storage. ## Create an InfluxDB Backup To create an InfluxDB backup, run the this command: `curl -k /platform/amv/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz", "nfs-server": "${fileSystemDns}", "nfs-path": "/"}'` ## Perform an InfluxDB Restore To perform an InfluxDB restore, run this command: `curl -k /platform/amv/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz", "nfs-server": "${fileSystemDns}", "nfs-path": "/"}'` The backup file is stored on the NFS storage. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) - [Installing Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) - [Upgrade Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/4.1/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring Backup and Restore for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) requires you to install, configure, and manage an NFS server, ensuring frequent backups are configured and performed. To preserve your data: - Enable automatic backups to protect against system failures or operational issues. - Schedule backups at least daily, or more frequently based on your needs. - Store backups externally, outside the Anypoint Platform cluster. The backup and restore process is controlled via Anypoint Platform PCE APIs. ## Create a Backup To create a backup, run the following command: `curl -k /platform/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}'` ### Backup Content
InformationIncluded in Anypoint Platform PCE Backup

Databases

Yes

Configurations

Yes

Secrets

Yes

Object Store

Yes

Disclaimer settings

Yes

SMTP settings

Yes

DNS settings

No

Platform certificate

No

NFS settings

No

NFS files

No

Drives

No

Runtimes*

No

Anypoint Monitoring metrics

No

\*Anypoint Platform PCE has information only about runtimes, such as their names and status. ## Perform a System Restore Consider the following before you perform a system restore: - Before running a restore, configure certain settings, such as NFS, certificates, and DNS because they are not included in the backup. - Backups cannot be transferred between different versions of Anypoint Platform PCE; therefore, your new installation must match the version of your backup. To perform the restore, run the following command: `curl -k /platform/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}` The backup file is stored on the NFS storage. ## Create an InfluxDB Backup To create an InfluxDB backup, run the this command: `curl -k /platform/amv/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}'` ## Perform an InfluxDB Restore To perform an InfluxDB restore, run this command: `curl -k /platform/amv/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}'` The backup file is stored on the NFS storage. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) - [Installing Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) - [Upgrade Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/4.2/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring Backup and Restore for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) stores backup archives on the platform shared filesystem. Depending on the storage mode you selected at installation, this is either an NFS server (NFS mode) or a dynamically provisioned volume from your RWX-capable `StorageClass` (StorageClass mode, Anypoint Platform PCE 4.3.0 and later). Make sure that the storage backing your shared filesystem is highly available and that frequent backups are configured and performed. > [!IMPORTANT] In StorageClass mode, use a `StorageClass` with `reclaimPolicy: Retain` for the > backup-restore volume so that deleting the `PersistentVolumeClaim` does not destroy historical > data. To preserve your data: - Enable automatic backups to protect against system failures or operational issues. - Schedule backups at least daily, or more frequently based on your needs. - Store backups externally, outside the Anypoint Platform cluster. The backup and restore process is controlled via Anypoint Platform PCE APIs. ## Create a Backup To create a backup, run the following command: `curl -k /platform/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}'` ### Backup Content
InformationIncluded in Anypoint Platform PCE Backup

Databases

Yes

Configurations

Yes

Secrets

Yes

Object Store

Yes

Disclaimer settings

Yes

SMTP settings

Yes

DNS settings

No

Platform certificate

No

Shared filesystem settings (NFS or StorageClass)

No

Shared filesystem contents

No

Drives

No

Runtimes*

No

Anypoint Monitoring metrics

No

\*Anypoint Platform PCE has information only about runtimes, such as their names and status. ## Perform a System Restore Consider the following before you perform a system restore: - Before running a restore, configure certain settings, such as the shared filesystem (NFS or StorageClass), certificates, and DNS because they are not included in the backup. - Backups cannot be transferred between different versions of Anypoint Platform PCE; therefore, your new installation must match the version of your backup. - The new installation must use the same storage mode (NFS or StorageClass) as the source installation. The installer rejects changes to the storage mode on an existing cluster. To perform the restore, run the following command: `curl -k /platform/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}` The backup file is stored on the platform shared filesystem. ## Create an InfluxDB Backup To create an InfluxDB backup, run the this command: `curl -k /platform/amv/backup-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}'` ## Perform an InfluxDB Restore To perform an InfluxDB restore, run this command: `curl -k /platform/amv/restore-X POST -H 'Authorization: Bearer ${token}' -H 'Content-Type: application/json' -d '{"backup-file-name": "${backupFileName}.tar.gz"}'` The backup file is stored on the platform shared filesystem. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) - [Installing Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/latest/install-workflow) - [Upgrade Anypoint Platform Private Cloud](https://docs.mulesoft.com/private-cloud/latest/upgrade) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring DNS on Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt For Mule runtime engine (Mule) to communicate with Anypoint Platform, you must provide the DNS or IP address where your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation is exposed to the local network. To do so, follow these steps: 1. From Anypoint Platform, click **Access Management**. 2. Click **DNS/IP**. 3. In **Platform DNS/IP**, enter the address where you are hosting the platform. 4. Click **Save**. > [!NOTE] If you perform a DNS update that changes the endpoints that Mule instances use to connect > to Runtime Manager, you must restart the Mule instances before they can connect to the new > endpoints. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring DNS on Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt For Mule runtime engine (Mule) to communicate with Anypoint Platform, you must provide the DNS or IP address where your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation is exposed to the local network. To do so, follow these steps: 1. From Anypoint Platform, click **Access Management**. 2. Click **DNS/IP**. 3. In **Platform DNS/IP**, enter the address where you are hosting the platform. 4. Click **Save**. > [!NOTE] If you perform a DNS update that changes the endpoints that Mule instances use to connect > to Runtime Manager, you must restart the Mule instances before they can connect to the new > endpoints. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring DNS on Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt For Mule runtime engine (Mule) to communicate with Anypoint Platform, you must provide the DNS or IP address where your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation is exposed to the local network. To do so, follow these steps: 1. From Anypoint Platform, click **Access Management**. 2. Click **DNS/IP**. 3. In **Platform DNS/IP**, enter the address where you are hosting the platform. 4. Click **Save**. > [!NOTE] If you perform a DNS update that changes the endpoints that Mule instances use to connect > to Runtime Manager, you must restart the Mule instances before they can connect to the new > endpoints. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring DNS on Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt For Mule runtime engine (Mule) to communicate with Anypoint Platform, you must provide the DNS or IP address where your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation is exposed to the local network. To do so, follow these steps: 1. From Anypoint Platform, click **Access Management**. 2. Click **DNS/IP**. 3. In **Platform DNS/IP**, enter the address where you are hosting the platform. 4. Click **Save**. > [!NOTE] If you perform a DNS update that changes the endpoints that Mule instances use to connect > to Runtime Manager, you must restart the Mule instances before they can connect to the new > endpoints. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring Grafana Dashboards > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Monitor the health of your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation and its running applications with Grafana dashboards. These dashboards provide a visual representation of the data collected by the monitoring tools in your Anypoint Platform PCE installation. The Grafana application uses a Kubernetes service of type `NodePort` to expose the service to the external network. To access Grafana, configure the load balancer to forward traffic to port `32380`. Alternatively, use the `kubectl port-forward` command to access Grafana locally: ```bash kubectl -n monitoring port-forward $(kubectl -n monitoring get svc --selector='app.kubernetes.io/name=grafana' -oname | head -n 1) 3000:80 ``` ## Access the Grafana Dashboard To retrieve credentials for Grafana, use these commands: ```bash kubectl -n monitoring get secrets grafana-creds -ojsonpath='{.data.admin-user}' | base64 --decode kubectl -n monitoring get secrets grafana-creds -ojsonpath='{.data.admin-password}' | base64 --decode ``` ## Create a Custom Dashboard You can create custom dashboards in Grafana using its web interface. However, these dashboards are lost if the Grafana pod restarts. To create a custom dashboard that persists across restarts, create a ConfigMap with the label `grafana_dashboard=1`. This example demonstrates how to create a ConfigMap with a custom dashboard: 1. Create a file with the dashboard configuration, for example, `example-redis.yaml`: ```yaml apiVersion: v1 kind: ConfigMap metadata: name: my-custom-dashboard namespace: monitoring labels: grafana_dashboard: "1" data: example-redis.json: | { "annotations": { "list": [ { "builtIn": 1, "datasource": "-- Grafana --", "enable": true, "hide": true, "iconColor": "rgba(0, 211, 255, 1)", "name": "Annotations & Alerts", "type": "dashboard" } ] }, "timezone": "browser", "title": "My Example Redis Dashboard", "description": "", "editable": true, "gnetId": null, "graphTooltip": 0, "id": 29, "iteration": 1645573265624, "links": [], "panels": [ { "aliasColors": { "Max": "#890F02", "Used": "#7EB26D", "max": "#BF1B00" }, "bars": false, "dashes": false, "datasource": null, "editable": false, "error": false, "fieldConfig": { "defaults": {}, "overrides": [] }, "fill": 1, "fillGradient": 0, "hiddenSeries": false, "id": 7, "isNew": true, "legend": { "avg": false, "current": false, "hideEmpty": false, "hideZero": false, "max": false, "min": false, "show": true, "total": false, "values": false }, "gridPos": { "x": 0, "y": 0, "h": 11, "w": 24 }, "lines": true, "linewidth": 2, "links": [], "nullPointMode": "null as zero", "options": { "alertThreshold": true }, "percentage": false, "pluginVersion": "7.5.10", "pointradius": 5, "points": false, "renderer": "flot", "seriesOverrides": [], "spaceLength": 10, "stack": false, "steppedLine": false, "targets": [ { "exemplar": true, "expr": "sum by (pod)(redis_used_memory{pod=~\"$pod\"})", "interval": "", "legendFormat": "{{pod}}", "refId": "B" } ], "thresholds": [], "timeFrom": null, "timeRegions": [], "timeShift": null, "title": "Memory Usage", "tooltip": { "msResolution": false, "shared": true, "sort": 0, "value_type": "cumulative" }, "type": "graph", "xaxis": { "buckets": null, "mode": "time", "name": null, "show": true, "values": [] }, "yaxes": [ { "format": "bytes", "label": "", "logBase": 1, "max": null, "min": 0, "show": true }, { "format": "short", "label": null, "logBase": 1, "max": null, "min": null, "show": false } ], "yaxis": { "align": false, "alignLevel": null } } ], "refresh": false, "schemaVersion": 27, "style": "dark", "tags": [], "templating": { "list": [ { "allValue": null, "current": { "selected": false, "text": "All", "value": "$__all" }, "datasource": null, "definition": "label_values(redis_maxmemory, pod)", "description": null, "error": null, "hide": 0, "includeAll": true, "label": "Pod", "multi": true, "name": "pod", "options": [], "query": { "query": "label_values(redis_maxmemory, pod)", "refId": "StandardVariableQuery" }, "refresh": 2, "regex": "", "skipUrlSync": false, "sort": 0, "tagValuesQuery": "", "tags": [], "tagsQuery": "", "type": "query", "useTags": false } ] }, "time": { "from": "now-6h", "to": "now" }, "timepicker": { "refresh_intervals": [ "5s", "10s", "30s", "1m", "5m", "15m", "30m", "1h", "2h", "1d" ], "time_options": [ "5m", "15m", "1h", "6h", "12h", "24h", "2d", "7d", "30d" ] } } ``` 2. Create the ConfigMap with the following command: ```bash kubectl apply -f example-redis.yaml ``` To update the dashboard, modify the ConfigMap, and the dashboard updates automatically. ## Delete a Custom Dashboard To delete a custom dashboard, delete the corresponding ConfigMap. The dashboard associated with the ConfigMap is removed automatically. ```bash kubectl -n monitoring delete cm my-custom-dashboard ``` ## List All Persistent Dashboards To get a list of all ConfigMaps with the label `grafana_dashboard=1`, use this command: kubectl get configmaps --selector='grafana\_dashboard=1' -A ## See Also - [Grafana documentation](https://grafana.com/docs/grafana/latest/visualizations/dashboards/) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring Grafana Dashboards > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Monitor the health of your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation and its running applications with Grafana dashboards. These dashboards provide a visual representation of the data collected by the monitoring tools in your Anypoint Platform PCE installation. The Grafana application uses a Kubernetes service of type `NodePort` to expose the service to the external network. To access Grafana, configure the load balancer to forward traffic to port `32380`. Alternatively, use the `kubectl port-forward` command to access Grafana locally: ```bash kubectl -n monitoring port-forward $(kubectl -n monitoring get svc --selector='app.kubernetes.io/name=grafana' -oname | head -n 1) 3000:80 ``` ## Access the Grafana Dashboard To retrieve credentials for Grafana, use these commands: ```bash kubectl -n monitoring get secrets grafana-creds -ojsonpath='{.data.admin-user}' | base64 --decode kubectl -n monitoring get secrets grafana-creds -ojsonpath='{.data.admin-password}' | base64 --decode ``` ## Create a Custom Dashboard You can create custom dashboards in Grafana using its web interface. However, these dashboards are lost if the Grafana pod restarts. To create a custom dashboard that persists across restarts, create a ConfigMap with the label `grafana_dashboard=1`. This example demonstrates how to create a ConfigMap with a custom dashboard: 1. Create a file with the dashboard configuration, for example, `example-redis.yaml`: ```yaml apiVersion: v1 kind: ConfigMap metadata: name: my-custom-dashboard namespace: monitoring labels: grafana_dashboard: "1" data: example-redis.json: | { "annotations": { "list": [ { "builtIn": 1, "datasource": "-- Grafana --", "enable": true, "hide": true, "iconColor": "rgba(0, 211, 255, 1)", "name": "Annotations & Alerts", "type": "dashboard" } ] }, "timezone": "browser", "title": "My Example Redis Dashboard", "description": "", "editable": true, "gnetId": null, "graphTooltip": 0, "id": 29, "iteration": 1645573265624, "links": [], "panels": [ { "aliasColors": { "Max": "#890F02", "Used": "#7EB26D", "max": "#BF1B00" }, "bars": false, "dashes": false, "datasource": null, "editable": false, "error": false, "fieldConfig": { "defaults": {}, "overrides": [] }, "fill": 1, "fillGradient": 0, "hiddenSeries": false, "id": 7, "isNew": true, "legend": { "avg": false, "current": false, "hideEmpty": false, "hideZero": false, "max": false, "min": false, "show": true, "total": false, "values": false }, "gridPos": { "x": 0, "y": 0, "h": 11, "w": 24 }, "lines": true, "linewidth": 2, "links": [], "nullPointMode": "null as zero", "options": { "alertThreshold": true }, "percentage": false, "pluginVersion": "7.5.10", "pointradius": 5, "points": false, "renderer": "flot", "seriesOverrides": [], "spaceLength": 10, "stack": false, "steppedLine": false, "targets": [ { "exemplar": true, "expr": "sum by (pod)(redis_used_memory{pod=~\"$pod\"})", "interval": "", "legendFormat": "{{pod}}", "refId": "B" } ], "thresholds": [], "timeFrom": null, "timeRegions": [], "timeShift": null, "title": "Memory Usage", "tooltip": { "msResolution": false, "shared": true, "sort": 0, "value_type": "cumulative" }, "type": "graph", "xaxis": { "buckets": null, "mode": "time", "name": null, "show": true, "values": [] }, "yaxes": [ { "format": "bytes", "label": "", "logBase": 1, "max": null, "min": 0, "show": true }, { "format": "short", "label": null, "logBase": 1, "max": null, "min": null, "show": false } ], "yaxis": { "align": false, "alignLevel": null } } ], "refresh": false, "schemaVersion": 27, "style": "dark", "tags": [], "templating": { "list": [ { "allValue": null, "current": { "selected": false, "text": "All", "value": "$__all" }, "datasource": null, "definition": "label_values(redis_maxmemory, pod)", "description": null, "error": null, "hide": 0, "includeAll": true, "label": "Pod", "multi": true, "name": "pod", "options": [], "query": { "query": "label_values(redis_maxmemory, pod)", "refId": "StandardVariableQuery" }, "refresh": 2, "regex": "", "skipUrlSync": false, "sort": 0, "tagValuesQuery": "", "tags": [], "tagsQuery": "", "type": "query", "useTags": false } ] }, "time": { "from": "now-6h", "to": "now" }, "timepicker": { "refresh_intervals": [ "5s", "10s", "30s", "1m", "5m", "15m", "30m", "1h", "2h", "1d" ], "time_options": [ "5m", "15m", "1h", "6h", "12h", "24h", "2d", "7d", "30d" ] } } ``` 2. Create the ConfigMap with the following command: ```bash kubectl apply -f example-redis.yaml ``` To update the dashboard, modify the ConfigMap, and the dashboard updates automatically. ## Delete a Custom Dashboard To delete a custom dashboard, delete the corresponding ConfigMap. The dashboard associated with the ConfigMap is removed automatically. ```bash kubectl -n monitoring delete cm my-custom-dashboard ``` ## List All Persistent Dashboards To get a list of all ConfigMaps with the label `grafana_dashboard=1`, use this command: kubectl get configmaps --selector='grafana\_dashboard=1' -A ## See Also - [Grafana documentation](https://grafana.com/docs/grafana/latest/visualizations/dashboards/) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring Grafana Dashboards > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Monitor the health of your Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) installation and its running applications with Grafana dashboards. These dashboards provide a visual representation of the data collected by the monitoring tools in your Anypoint Platform PCE installation. The Grafana application uses a Kubernetes service of type `NodePort` to expose the service to the external network. To access Grafana, configure the load balancer to forward traffic to port `32380`. Alternatively, use the `kubectl port-forward` command to access Grafana locally: ```bash kubectl -n monitoring port-forward $(kubectl -n monitoring get svc --selector='app.kubernetes.io/name=grafana' -oname | head -n 1) 3000:80 ``` ## Access the Grafana Dashboard To retrieve credentials for Grafana, use these commands: ```bash kubectl -n monitoring get secrets grafana-creds -ojsonpath='{.data.admin-user}' | base64 --decode kubectl -n monitoring get secrets grafana-creds -ojsonpath='{.data.admin-password}' | base64 --decode ``` ## Create a Custom Dashboard You can create custom dashboards in Grafana using its web interface. However, these dashboards are lost if the Grafana pod restarts. To create a custom dashboard that persists across restarts, create a ConfigMap with the label `grafana_dashboard=1`. This example demonstrates how to create a ConfigMap with a custom dashboard: 1. Create a file with the dashboard configuration, for example, `example-redis.yaml`: ```yaml apiVersion: v1 kind: ConfigMap metadata: name: my-custom-dashboard namespace: monitoring labels: grafana_dashboard: "1" data: example-redis.json: | { "annotations": { "list": [ { "builtIn": 1, "datasource": "-- Grafana --", "enable": true, "hide": true, "iconColor": "rgba(0, 211, 255, 1)", "name": "Annotations & Alerts", "type": "dashboard" } ] }, "timezone": "browser", "title": "My Example Redis Dashboard", "description": "", "editable": true, "gnetId": null, "graphTooltip": 0, "id": 29, "iteration": 1645573265624, "links": [], "panels": [ { "aliasColors": { "Max": "#890F02", "Used": "#7EB26D", "max": "#BF1B00" }, "bars": false, "dashes": false, "datasource": null, "editable": false, "error": false, "fieldConfig": { "defaults": {}, "overrides": [] }, "fill": 1, "fillGradient": 0, "hiddenSeries": false, "id": 7, "isNew": true, "legend": { "avg": false, "current": false, "hideEmpty": false, "hideZero": false, "max": false, "min": false, "show": true, "total": false, "values": false }, "gridPos": { "x": 0, "y": 0, "h": 11, "w": 24 }, "lines": true, "linewidth": 2, "links": [], "nullPointMode": "null as zero", "options": { "alertThreshold": true }, "percentage": false, "pluginVersion": "7.5.10", "pointradius": 5, "points": false, "renderer": "flot", "seriesOverrides": [], "spaceLength": 10, "stack": false, "steppedLine": false, "targets": [ { "exemplar": true, "expr": "sum by (pod)(redis_used_memory{pod=~\"$pod\"})", "interval": "", "legendFormat": "{{pod}}", "refId": "B" } ], "thresholds": [], "timeFrom": null, "timeRegions": [], "timeShift": null, "title": "Memory Usage", "tooltip": { "msResolution": false, "shared": true, "sort": 0, "value_type": "cumulative" }, "type": "graph", "xaxis": { "buckets": null, "mode": "time", "name": null, "show": true, "values": [] }, "yaxes": [ { "format": "bytes", "label": "", "logBase": 1, "max": null, "min": 0, "show": true }, { "format": "short", "label": null, "logBase": 1, "max": null, "min": null, "show": false } ], "yaxis": { "align": false, "alignLevel": null } } ], "refresh": false, "schemaVersion": 27, "style": "dark", "tags": [], "templating": { "list": [ { "allValue": null, "current": { "selected": false, "text": "All", "value": "$__all" }, "datasource": null, "definition": "label_values(redis_maxmemory, pod)", "description": null, "error": null, "hide": 0, "includeAll": true, "label": "Pod", "multi": true, "name": "pod", "options": [], "query": { "query": "label_values(redis_maxmemory, pod)", "refId": "StandardVariableQuery" }, "refresh": 2, "regex": "", "skipUrlSync": false, "sort": 0, "tagValuesQuery": "", "tags": [], "tagsQuery": "", "type": "query", "useTags": false } ] }, "time": { "from": "now-6h", "to": "now" }, "timepicker": { "refresh_intervals": [ "5s", "10s", "30s", "1m", "5m", "15m", "30m", "1h", "2h", "1d" ], "time_options": [ "5m", "15m", "1h", "6h", "12h", "24h", "2d", "7d", "30d" ] } } ``` 2. Create the ConfigMap with the following command: ```bash kubectl apply -f example-redis.yaml ``` To update the dashboard, modify the ConfigMap, and the dashboard updates automatically. ## Delete a Custom Dashboard To delete a custom dashboard, delete the corresponding ConfigMap. The dashboard associated with the ConfigMap is removed automatically. ```bash kubectl -n monitoring delete cm my-custom-dashboard ``` ## List All Persistent Dashboards To get a list of all ConfigMaps with the label `grafana_dashboard=1`, use this command: kubectl get configmaps --selector='grafana\_dashboard=1' -A ## See Also - [Grafana documentation](https://grafana.com/docs/grafana/latest/visualizations/dashboards/) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring Security for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To use Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), you must supply TLS credentials. Upload a certificate via the Anypoint Platform UI under **Access Management**, and configure the platform to use this certificate to secure communications. Ensure that every machine connected to the platform, including Mule runtime engine (Mule), trusts this certificate. Additionally, you have the option to specify custom ciphers and Diffie-Hellman (DH) groups for the key exchange method if needed. When you replace or change the certificate, Anypoint Platform restarts the nginx server. ![Certificate and Private Key configuration under the Security tab in Access Management](/mulesoft/current/en-us/private-cloud/4.0/_images/on-prem-features-5f98d.png) To configure the security, follow these steps: 1. Create a certificate. Generate your certificate using a tool like OpenSSL. To be compatible with Anypoint Platform, ensure your certificate is in an `nginx` friendly format. The certificate must include the domain name of your installation as both the Common Name (CN) and the Subject Alternative Name (SAN). 2. From Anypoint Platform, select **Access Management**. 3. Select **Security**. 4. Under **Certificate**, select **Browse** to select the certificate on your local system. 5. Under **Key**, select **Browse** to select the private key for your certificate. 6. For additional security, you can upload your own cipher and Diffie-Hellman group file. Note that this step is optional. 1. Deselect **Use default ciphers and DH Group**. 2. Enter a set of ciphers from the list of valid NGINX ciphers. If you are using Open SSL, you can view this list by running the `openssl ciphers` command. The cipher you enter must be in an `nginx` compatible format, for example: ECDHE-RSA-AES256-GCM-SHA512:DHE-RSA-AES256-GCM-SHA512:ECDHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-SHA384 3. Generate `dhparam.pem` (the Diffie-Hellman group file) using the following `openssl` command: openssl dhparam 2048 -out dhparam.pem 4. Upload the `dhparam.pem` file from your system. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring Security for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To use Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), you must supply TLS credentials. Upload a certificate via the Anypoint Platform UI under **Access Management**, and configure the platform to use this certificate to secure communications. Ensure that every machine connected to the platform, including Mule runtime engine (Mule), trusts this certificate. Additionally, you have the option to specify custom ciphers and Diffie-Hellman (DH) groups for the key exchange method if needed. When you replace or change the certificate, Anypoint Platform restarts the nginx server. ![Certificate and Private Key configuration under the Security tab in Access Management](/mulesoft/current/en-us/private-cloud/4.1/_images/on-prem-features-5f98d.png) To configure the security, follow these steps: 1. Create a certificate. Generate your certificate using a tool like OpenSSL. To be compatible with Anypoint Platform, ensure your certificate is in an `nginx` friendly format. The certificate must include the domain name of your installation as both the Common Name (CN) and the Subject Alternative Name (SAN). 2. From Anypoint Platform, select **Access Management**. 3. Select **Security**. 4. Under **Certificate**, select **Browse** to select the certificate on your local system. 5. Under **Key**, select **Browse** to select the private key for your certificate. 6. For additional security, you can upload your own cipher and Diffie-Hellman group file. Note that this step is optional. 1. Deselect **Use default ciphers and DH Group**. 2. Enter a set of ciphers from the list of valid NGINX ciphers. If you are using Open SSL, you can view this list by running the `openssl ciphers` command. The cipher you enter must be in an `nginx` compatible format, for example: ECDHE-RSA-AES256-GCM-SHA512:DHE-RSA-AES256-GCM-SHA512:ECDHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-SHA384 3. Generate `dhparam.pem` (the Diffie-Hellman group file) using the following `openssl` command: openssl dhparam 2048 -out dhparam.pem 4. Upload the `dhparam.pem` file from your system. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring Security for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To use Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), you must supply TLS credentials. Upload a certificate via the Anypoint Platform UI under **Access Management**, and configure the platform to use this certificate to secure communications. Ensure that every machine connected to the platform, including Mule runtime engine (Mule), trusts this certificate. Additionally, you have the option to specify custom ciphers and Diffie-Hellman (DH) groups for the key exchange method if needed. When you replace or change the certificate, Anypoint Platform restarts the nginx server. ![Certificate and Private Key configuration under the Security tab in Access Management](/mulesoft/current/en-us/private-cloud/4.2/_images/on-prem-features-5f98d.png) To configure the security, follow these steps: 1. Create a certificate. Generate your certificate using a tool like OpenSSL. To be compatible with Anypoint Platform, ensure your certificate is in an `nginx` friendly format. The certificate must include the domain name of your installation as both the Common Name (CN) and the Subject Alternative Name (SAN). 2. From Anypoint Platform, select **Access Management**. 3. Select **Security**. 4. Under **Certificate**, select **Browse** to select the certificate on your local system. 5. Under **Key**, select **Browse** to select the private key for your certificate. 6. For additional security, you can upload your own cipher and Diffie-Hellman group file. Note that this step is optional. 1. Deselect **Use default ciphers and DH Group**. 2. Enter a set of ciphers from the list of valid NGINX ciphers. If you are using Open SSL, you can view this list by running the `openssl ciphers` command. The cipher you enter must be in an `nginx` compatible format, for example: ECDHE-RSA-AES256-GCM-SHA512:DHE-RSA-AES256-GCM-SHA512:ECDHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-SHA384 3. Generate `dhparam.pem` (the Diffie-Hellman group file) using the following `openssl` command: openssl dhparam 2048 -out dhparam.pem 4. Upload the `dhparam.pem` file from your system. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring Security for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To use Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), you must supply TLS credentials. Upload a certificate via the Anypoint Platform UI under **Access Management**, and configure the platform to use this certificate to secure communications. Ensure that every machine connected to the platform, including Mule runtime engine (Mule), trusts this certificate. Additionally, you have the option to specify custom ciphers and Diffie-Hellman (DH) groups for the key exchange method if needed. When you replace or change the certificate, Anypoint Platform restarts the nginx server. ![Certificate and Private Key configuration under the Security tab in Access Management](/mulesoft/current/en-us/private-cloud/latest/_images/on-prem-features-5f98d.png) To configure the security, follow these steps: 1. Create a certificate. Generate your certificate using a tool like OpenSSL. To be compatible with Anypoint Platform, ensure your certificate is in an `nginx` friendly format. The certificate must include the domain name of your installation as both the Common Name (CN) and the Subject Alternative Name (SAN). 2. From Anypoint Platform, select **Access Management**. 3. Select **Security**. 4. Under **Certificate**, select **Browse** to select the certificate on your local system. 5. Under **Key**, select **Browse** to select the private key for your certificate. 6. For additional security, you can upload your own cipher and Diffie-Hellman group file. Note that this step is optional. 1. Deselect **Use default ciphers and DH Group**. 2. Enter a set of ciphers from the list of valid NGINX ciphers. If you are using Open SSL, you can view this list by running the `openssl ciphers` command. The cipher you enter must be in an `nginx` compatible format, for example: ECDHE-RSA-AES256-GCM-SHA512:DHE-RSA-AES256-GCM-SHA512:ECDHE-RSA-AES256-GCM-SHA384:DHE-RSA-AES256-GCM-SHA384:ECDHE-RSA-AES256-SHA384 3. Generate `dhparam.pem` (the Diffie-Hellman group file) using the following `openssl` command: openssl dhparam 2048 -out dhparam.pem 4. Upload the `dhparam.pem` file from your system. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Configuring SMTP in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Because Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) does not provide an SMTP server, you must set up your own SMTP email server to deliver email alerts generated by Runtime Manager. The SMTP email server also sends email messages for platform infrastructure alerts, including CPU, memory, and disk alerts. You must direct your local installation of Anypoint Platform PCE to this server from the Access Management panel. > [!NOTE] Only SMTP servers with authentication are supported (anonymous SMTP relay servers are not > supported). SMTPS is not supported. ## Connect to an Existing SMTP Server 1. From Anypoint Platform, select **Access Management**. 2. Select **SMTP**. 3. Provide the following parameters information for your SMTP server:
FieldDescriptionExample

Host

Host name of your SMTP server.

smtp.gmail.com

Port

Port used to communicate to your SMTP server.

587

Username

Username to authenticate to your SMTP server.

mulesoft.onprem.smtp@gmail.com

Password

SMTP password.

examplepassphrase

> [!NOTE] To use anonymous logins for your SMTP server, leave the **Username** and **Password** > fields empty. 4. After entering your configuration, select **Save**. ![SMTP configuration window under the Access Management UI](/mulesoft/current/en-us/private-cloud/4.0/_images/on-premises-features-2c60a.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Configuring SMTP in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Because Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) does not provide an SMTP server, you must set up your own SMTP email server to deliver email alerts generated by Runtime Manager. The SMTP email server also sends email messages for platform infrastructure alerts, including CPU, memory, and disk alerts. You must direct your local installation of Anypoint Platform PCE to this server from the Access Management panel. > [!NOTE] Only SMTP servers with authentication are supported (anonymous SMTP relay servers are not > supported). SMTPS is not supported. ## Connect to an Existing SMTP Server 1. From Anypoint Platform, select **Access Management**. 2. Select **SMTP**. 3. Provide the following parameters information for your SMTP server:
FieldDescriptionExample

Host

Host name of your SMTP server.

smtp.gmail.com

Port

Port used to communicate to your SMTP server.

587

Username

Username to authenticate to your SMTP server.

mulesoft.onprem.smtp@gmail.com

Password

SMTP password.

examplepassphrase

> [!NOTE] To use anonymous logins for your SMTP server, leave the **Username** and **Password** > fields empty. 4. After entering your configuration, select **Save**. ![SMTP configuration window under the Access Management UI](/mulesoft/current/en-us/private-cloud/4.1/_images/on-premises-features-2c60a.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Configuring SMTP in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Because Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) does not provide an SMTP server, you must set up your own SMTP email server to deliver email alerts generated by Runtime Manager. The SMTP email server also sends email messages for platform infrastructure alerts, including CPU, memory, and disk alerts. You must direct your local installation of Anypoint Platform PCE to this server from the Access Management panel. > [!NOTE] Only SMTP servers with authentication are supported (anonymous SMTP relay servers are not > supported). SMTPS is not supported. ## Connect to an Existing SMTP Server 1. From Anypoint Platform, select **Access Management**. 2. Select **SMTP**. 3. Provide the following parameters information for your SMTP server:
FieldDescriptionExample

Host

Host name of your SMTP server.

smtp.gmail.com

Port

Port used to communicate to your SMTP server.

587

Username

Username to authenticate to your SMTP server.

mulesoft.onprem.smtp@gmail.com

Password

SMTP password.

examplepassphrase

> [!NOTE] To use anonymous logins for your SMTP server, leave the **Username** and **Password** > fields empty. 4. After entering your configuration, select **Save**. ![SMTP configuration window under the Access Management UI](/mulesoft/current/en-us/private-cloud/4.2/_images/on-premises-features-2c60a.png) ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Configuring SMTP in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Because Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) does not provide an SMTP server, you must set up your own SMTP email server to deliver email alerts generated by Runtime Manager. The SMTP email server also sends email messages for platform infrastructure alerts, including CPU, memory, and disk alerts. You must direct your local installation of Anypoint Platform PCE to this server from the Access Management panel. > [!NOTE] Only SMTP servers with authentication are supported (anonymous SMTP relay servers are not > supported). SMTPS is not supported. ## Connect to an Existing SMTP Server 1. From Anypoint Platform, select **Access Management**. 2. Select **SMTP**. 3. Provide the following parameters information for your SMTP server:
FieldDescriptionExample

Host

Host name of your SMTP server.

smtp.gmail.com

Port

Port used to communicate to your SMTP server.

587

Username

Username to authenticate to your SMTP server.

mulesoft.onprem.smtp@gmail.com

Password

SMTP password.

examplepassphrase

Sender

Optional. Email address that alert notifications are sent from. Available starting with Anypoint Platform PCE 4.3.0. Leave this field blank to use the default sender address.

alerts@example.com

> [!NOTE] To use anonymous logins for your SMTP server, leave the **Username** and **Password** > fields empty. 4. After entering your configuration, select **Save**. ![SMTP configuration window under the Access Management UI](/mulesoft/current/en-us/private-cloud/latest/_images/on-premises-features-2c60a.png) ## Set a Custom Sender Address In Anypoint Platform PCE versions before 4.3.0, outbound alert emails use a fixed sender address. Mail systems that accept messages only from approved internal domains can reject those emails, and Anypoint Platform PCE doesn’t report the rejection. Starting with Anypoint Platform PCE 4.3.0, the SMTP page in **Access Management** includes an optional **Sender** field. To set the address: 1. From Anypoint Platform, select **Access Management**. 2. Select **SMTP**. 3. Configure the **Host**, **Port**, **Username**, and **Password** fields. 4. In **Sender**, enter the email address to send alert notifications from. 5. Select **Save**. Anypoint Platform PCE checks that **Sender** is a properly formatted email address. If you leave **Sender** blank, Anypoint Platform PCE uses the default sender behavior from versions before 4.3.0. > [!IMPORTANT] Anypoint Platform PCE checks only that **Sender** is a properly formatted email > address. It doesn’t check that your SMTP server is authorized to send from that address or domain. > Use an address on a domain your SMTP server is authorized to send from. If the address is outside > those domains, alert emails can fail to send, and Anypoint Platform PCE doesn’t show an error. > Confirm the domains your SMTP server allows before you set **Sender**. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Delete the Local User After Installing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt During installation, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) creates a local Anypoint Platform user that enables you to configure platform components during installation. The credentials for this user are stored locally in Anypoint Platform. However, after configuring an external authentication provider such as LDAP, you should delete this user to improve security. 1. Before deleting the local user, perform the following: - Install and configure your external identity provider. - Ensure that `SystemAdmin` user is not the owner of any other organizations. 2. Select the new user to whom you want to assign administrator privileges. This user must be defined in your external identity provider. 3. Determine the internal user ID for this user. 1. From Anypoint Platform, select Access Management. 2. Select Users and then select the name of the user. 3. Record the internal user ID. This ID appears in the URL as: ![User ID](/mulesoft/current/en-us/private-cloud/3.2/_images/access-management-user-id.png) 4. Enter the gravity shell: gravity enter 5. Identify one of the `authentication-server` containers: kubectl get pods -n access-management -l component=authentication-server 6. Change the owner, providing the internal the `authentication-server` container and user ID obtained earlier: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id This command changes the owner from the default system user to a new user managed by the external authentication provider. The command you use should be similar to the following example: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id a363279f-982f-493c-b08f-9feb91be90d4 7. Disable and delete the `System Admin` user. 1. Log in as the user you just added as the organization owner. 2. From Anypoint Platform, select Access Management. 3. Select Users and then select `username` in the same row as `SystemAdmin`. 4. Select Disable and then select Delete. This removes the default user from the platform. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Deleting the Local User After Installing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt During installation, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) creates a local Anypoint Platform user that enables you to configure platform components during installation. The credentials for this user are stored locally in Anypoint Platform. However, after configuring an external authentication provider such as LDAP, you should delete this user to improve security. 1. Before deleting the local user, perform the following steps: - Install and configure your external identity provider. - Ensure that `SystemAdmin` user is not the owner of any other organizations. 2. Select the new user to whom you want to assign administrator privileges. This user must be defined in your external identity provider. 3. Determine the internal user ID for this user. 1. From Anypoint Platform, select **Access Management**. 2. Select **Users**. 3. Select the name of the user. 4. Record the internal user ID. This ID appears in the URL as: ![Example of User ID in the URL](/mulesoft/current/en-us/private-cloud/4.0/_images/access-management-user-id.png) 4. Configure your installer machine with \`kubeconfig.yaml to access the Anypoint Platform PCE K8s cluster. 1. Set the environment variable `KUBECONFIG=/path/to/kubeconfig.yaml`. 2. Check access to kubernetes cluster with `kubectl get nodes`. 5. Identify one of the `authentication-server` containers: kubectl get pods -n access-management -l component=authentication-server 6. Change the owner, providing the internal the `authentication-server` container and user ID obtained earlier: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id This command changes the owner from the default system user to a new user managed by the external authentication provider. The command you use should be similar to the following example: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id a363279f-982f-493c-b08f-9feb91be90d4 7. Disable and delete the `System Admin` user. 1. Log in as the user you just added as the organization owner. 2. From Anypoint Platform, select **Access Management**. 3. Select **Users**. 4. Select `username` in the same row as `SystemAdmin`. 5. Select **Disable**. 6. Select **Delete**. This removes the default user from the platform. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Deleting the Local User After Installing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt During installation, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) creates a local Anypoint Platform user that enables you to configure platform components during installation. The credentials for this user are stored locally in Anypoint Platform. However, after configuring an external authentication provider such as LDAP, you should delete this user to improve security. 1. Before deleting the local user, perform the following steps: - Install and configure your external identity provider. - Ensure that `SystemAdmin` user is not the owner of any other organizations. 2. Select the new user to whom you want to assign administrator privileges. This user must be defined in your external identity provider. 3. Determine the internal user ID for this user. 1. From Anypoint Platform, select **Access Management**. 2. Select **Users**. 3. Select the name of the user. 4. Record the internal user ID. This ID appears in the URL as: ![Example of User ID in the URL](/mulesoft/current/en-us/private-cloud/4.1/_images/access-management-user-id.png) 4. Configure your installer machine with \`kubeconfig.yaml to access the Anypoint Platform PCE K8s cluster. 1. Set the environment variable `KUBECONFIG=/path/to/kubeconfig.yaml`. 2. Check access to kubernetes cluster with `kubectl get nodes`. 5. Identify one of the `authentication-server` containers: kubectl get pods -n access-management -l component=authentication-server 6. Change the owner, providing the internal the `authentication-server` container and user ID obtained earlier: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id This command changes the owner from the default system user to a new user managed by the external authentication provider. The command you use should be similar to the following example: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id a363279f-982f-493c-b08f-9feb91be90d4 7. Disable and delete the `System Admin` user. 1. Log in as the user you just added as the organization owner. 2. From Anypoint Platform, select **Access Management**. 3. Select **Users**. 4. Select `username` in the same row as `SystemAdmin`. 5. Select **Disable**. 6. Select **Delete**. This removes the default user from the platform. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Deleting the Local User After Installing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt During installation, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) creates a local Anypoint Platform user that enables you to configure platform components during installation. The credentials for this user are stored locally in Anypoint Platform. However, after configuring an external authentication provider such as LDAP, you should delete this user to improve security. 1. Before deleting the local user, perform the following steps: - Install and configure your external identity provider. - Ensure that `SystemAdmin` user is not the owner of any other organizations. 2. Select the new user to whom you want to assign administrator privileges. This user must be defined in your external identity provider. 3. Determine the internal user ID for this user. 1. From Anypoint Platform, select **Access Management**. 2. Select **Users**. 3. Select the name of the user. 4. Record the internal user ID. This ID appears in the URL as: ![Example of User ID in the URL](/mulesoft/current/en-us/private-cloud/4.2/_images/access-management-user-id.png) 4. Configure your installer machine with \`kubeconfig.yaml to access the Anypoint Platform PCE K8s cluster. 1. Set the environment variable `KUBECONFIG=/path/to/kubeconfig.yaml`. 2. Check access to kubernetes cluster with `kubectl get nodes`. 5. Identify one of the `authentication-server` containers: kubectl get pods -n access-management -l component=authentication-server 6. Change the owner, providing the internal the `authentication-server` container and user ID obtained earlier: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id This command changes the owner from the default system user to a new user managed by the external authentication provider. The command you use should be similar to the following example: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id a363279f-982f-493c-b08f-9feb91be90d4 7. Disable and delete the `System Admin` user. 1. Log in as the user you just added as the organization owner. 2. From Anypoint Platform, select **Access Management**. 3. Select **Users**. 4. Select `username` in the same row as `SystemAdmin`. 5. Select **Disable**. 6. Select **Delete**. This removes the default user from the platform. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Deleting the Local User After Installing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt During installation, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) creates a local Anypoint Platform user that enables you to configure platform components during installation. The credentials for this user are stored locally in Anypoint Platform. However, after configuring an external authentication provider such as LDAP, you should delete this user to improve security. 1. Before deleting the local user, perform the following steps: - Install and configure your external identity provider. - Ensure that `SystemAdmin` user is not the owner of any other organizations. 2. Select the new user to whom you want to assign administrator privileges. This user must be defined in your external identity provider. 3. Determine the internal user ID for this user. 1. From Anypoint Platform, select **Access Management**. 2. Select **Users**. 3. Select the name of the user. 4. Record the internal user ID. This ID appears in the URL as: ![Example of User ID in the URL](/mulesoft/current/en-us/private-cloud/latest/_images/access-management-user-id.png) 4. Configure your installer machine with \`kubeconfig.yaml to access the Anypoint Platform PCE K8s cluster. 1. Set the environment variable `KUBECONFIG=/path/to/kubeconfig.yaml`. 2. Check access to kubernetes cluster with `kubectl get nodes`. 5. Identify one of the `authentication-server` containers: kubectl get pods -n access-management -l component=authentication-server 6. Change the owner, providing the internal the `authentication-server` container and user ID obtained earlier: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id This command changes the owner from the default system user to a new user managed by the external authentication provider. The command you use should be similar to the following example: kubectl exec -it -n access-management -c application – node bin/change\_owner.js --new-owner-id a363279f-982f-493c-b08f-9feb91be90d4 7. Disable and delete the `System Admin` user. 1. Log in as the user you just added as the organization owner. 2. From Anypoint Platform, select **Access Management**. 3. Select **Users**. 4. Select `username` in the same row as `SystemAdmin`. 5. Select **Disable**. 6. Select **Delete**. This removes the default user from the platform. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Deploying Mule Applications to Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and register your hybrid servers, you can deploy Mule applications to those servers through Runtime Manager. Before you deploy, ensure that: - Your servers are registered and running in Runtime Manager. See [About Servers, Server Groups, and Clusters](https://docs.mulesoft.com/runtime-manager/managing-servers). - Your IDE is configured to use your Anypoint Platform PCE control plane rather than the default cloud Anypoint Platform. You can develop and package applications with either Anypoint Code Builder or Anypoint Studio, then deploy them to Anypoint Platform PCE. ## Deploy Using Anypoint Code Builder Use Anypoint Code Builder in VS Code to develop, package, and deploy Mule applications to Anypoint Platform PCE. You can follow the standard Anypoint Code Builder documentation without MuleSoft Vibes. MuleSoft Vibes is optional. - [Configuring Anypoint Code Builder for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-code-builder) - [Deploying Mule Applications Using Anypoint Code Builder](https://docs.mulesoft.com/private-cloud/latest/deploy-apps-code-builder) ## Deploy Using Anypoint Studio Configure Anypoint Studio to connect to your Anypoint Platform PCE instance, then package your application and deploy it to a hybrid server in Runtime Manager: - [Configuring Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-studio) - [Export a Studio Project to a Deployable Mule Application](https://docs.mulesoft.com/studio/import-export-packages) - [Deploying to Your Own Servers](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers) > [!NOTE] Anypoint Studio’s **Deploy to CloudHub** option targets CloudHub and does not deploy > applications to Anypoint Platform PCE hybrid servers. For PCE, export a deployable JAR and upload > it in Runtime Manager. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) - [Deploying to Your Own Servers](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers) - [Troubleshooting Mule Runtime Engine Registration](https://docs.mulesoft.com/private-cloud/latest/register-server) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Deploying Mule Applications Using Anypoint Code Builder > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Follow these steps to develop, package, and deploy Mule applications to Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) using Anypoint Code Builder in VS Code. You can complete this workflow without MuleSoft Vibes by following the standard Anypoint Code Builder documentation. MuleSoft Vibes is optional. 1. [Connect Anypoint Platform PCE with Anypoint Code Builder](#connect-acb) 2. [Develop Mule Application](#develop-mule-app) 3. [Package and Export the Mule Application](#package-export) 4. [Deploy to Runtime Manager](#deploy-runtime-manager) 5. [Publish to Exchange](#publish-exchange) 6. [Validate the Deployment](#validate-deployment) ## Connect Anypoint Platform PCE with Anypoint Code Builder To connect Anypoint Platform PCE with Anypoint Code Builder, follow the instructions in [Configuring Anypoint Code Builder for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-code-builder). Private cloud mode connects Anypoint Code Builder on your machine to the Anypoint Platform PCE control plane. Sign-in uses OAuth or OIDC with the identity provider hosted by Anypoint Platform PCE. Anypoint Code Builder keeps application metadata in your network and blocks standard cloud telemetry. ## Develop Mule Application Build the Mule application in VS Code with Anypoint Code Builder. You can create the specification, the flow, DataWeave transformations, and security configuration without MuleSoft Vibes. Follow the standard Anypoint Code Builder documentation: - [Designing API Specs and Fragments](https://docs.mulesoft.com/anypoint-code-builder/des-designing-api-specs) - [Developing Integrations](https://docs.mulesoft.com/anypoint-code-builder/int-developing-integrations) MuleSoft Vibes is optional. If you use it, Vibes skills can generate the specification, the flow, DataWeave transformations, and security configuration. These skills are available in the skill pack: - `build-mule-integration` generates integration flows, configures endpoints, and scaffolds orchestration. - `secure-mule-app` scans the configuration for security risks, applies encryption, and suggests logging. - DataWeave skills generate transformations and map payload schemas. You can add organization-specific skills that enforce naming conventions, header standards, and logging policies. Package those skills in a repository and use them with Anypoint Code Builder. 1. In VS Code, open the Command Palette and run `MuleSoft: Design an API` to create the API specification. Anypoint Code Builder validates the specification locally. 2. To mock the API on your machine, use the API Console. 3. Run `MuleSoft: Implement an API Specification` to stub the interface and connect the services. 4. To create a project from the canvas instead, select the Anypoint Code Builder icon in the primary side bar, select `Develop an Integration`, and create a Mule project. 5. To import an asset from Exchange, run `MuleSoft: Import Asset From Exchange`. Select the asset type and add it to the project. 6. Develop the flow, DataWeave transformations, and security configuration in Anypoint Code Builder. If you use MuleSoft Vibes, you can use natural language prompts for flow logic, DataWeave mappings, and the `secure-mule-app` skill. 7. Create and run MUnit tests in the project before you package the application. 8. Run a local debug session from the **Run and Debug** sidebar to test the logic before packaging. ![Anypoint Code Builder in VS Code, with the command palette open to MuleSoft: Login to Anypoint Platform, a Mule flow on the canvas, and the local runtime in the terminal](/mulesoft/current/en-us/private-cloud/latest/_images/pce-acb-vibes.png) ## Package and Export the Mule Application Manually create the deployable Mule application artifact: 1. In the VS Code Explorer, right-click your `mule-artifact.json` file or any XML file in `src/main/mule`. 2. Select **MuleSoft: Export Mule Project**. 3. Choose **Deployable JAR** and select **Include project sources**. Including project sources enables you to re-import the project later if needed. 4. Save the JAR to your local machine. ## Deploy to Runtime Manager Upload the JAR to your hybrid server. The server must be registered and in the **Running** state. 1. Log in to Anypoint Platform PCE and navigate to **Runtime Manager**. 2. Select the correct environment, and then click **Deploy Application**. 3. Set the **Deployment Target** to your hybrid server or cluster. 4. Click **Choose File** > **Upload File** and select the JAR you exported. 5. Click **Deploy Application**. After the application status is **Started**, apply API Manager policies to the endpoints. Use API Manager in Anypoint Platform PCE, or the Anypoint CLI API Manager plugin from [Configuring Anypoint Code Builder for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-code-builder). ## Publish to Exchange Publishing an asset directly from Anypoint Code Builder to Exchange in Anypoint Platform PCE has partial support. Publish the deployable JAR from Exchange in Anypoint Platform PCE: 1. In Anypoint Platform PCE, navigate to **Exchange**. 2. Click **Publish New Asset**. 3. Provide an asset name and select **Mule Application** as the asset type. 4. Upload the same JAR file. This publishes the project to your organization’s private library. ## Validate the Deployment To validate the deployment, follow these steps: 1. In Runtime Manager, wait until the application status is **Started**. 2. Use an API client such as Postman or the VS Code Thunder Client to send a request to the server IP and the port defined in your HTTP Listener. For example: `[http://10.x.x.x:8081/api/test](http://10.x.x.x:8081/api/test)` 3. Check the **Logs** tab in Runtime Manager to verify that the request was processed successfully. ## See Also - [Deploying Mule Applications to Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/deploy-apps) - [Configuring Anypoint Code Builder for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-code-builder) - [Developing Integrations](https://docs.mulesoft.com/anypoint-code-builder/int-developing-integrations) - [Using Skills with MuleSoft Vibes](https://docs.mulesoft.com/anypoint-code-builder/vibes-skills) - [Deploying to Your Own Servers](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers) - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Environment Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt ## Supported Environments The following environments are the only environments supported by Anypoint Platform Private Cloud Edition (Anypoint Platform PCE): - VMWare bare metal virtualization - Physical servers - Amazon Web Services (AWS) If you are using AWS, perform the steps provided in [Provision AWS Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-aws-terraform). ## See Also - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Gather Environment Information > For the full documentation index, see: https://docs.mulesoft.com/llms.txt During installation, you will be asked to provide the following information: - Your Anypoint Platform Private Cloud Edition license file. For example, `license.pem`. Contact your MuleSoft representative for assistance. - Name and location of the installation package. For example, `/home/ec2-user/anypoint-2.0.3-installer.tar.gz`. Contact your MuleSoft representative for assistance. - AWS credentials - Admin ID - Admin email address - Admin password - IP address for each server in the cluster - Device names - System state directory - Application data - etcd - Docker - NFS server information - DNS for NFS server - NFS server path - PLATFORM\_DNS (The DNS must be configured to the NFS server.) - Anypoint Platform organization and administrator: - Administrator username - Administrator email - Administrator password - SMTP server credentials - Host name - Port used to communicate to your SMTP server - User name to authenticate to your SMTP server - SMTP password ## See Also - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Hardware Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To ensure the performance and stability of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), every node in your Anypoint Platform PCE environment must meet the requirements described in this topic. > [!IMPORTANT] Before you install Anypoint Platform PCE, your infrastructure team must review each > of the following sections and verify that your environment meets the stated disk and device > requirements. If needed, contact your MuleSoft representative for assistance. There are 2 components of PCE: - Anypoint Platform Core - Anypoint Monitoring (optional) To ensure platform performance, stability, and high availability, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) supports two network configurations: 4-node and 7-node. Small configuration: - 4 nodes for Control Plane - 3 additional nodes for Anypoint Monitoring (optional add-on) - NFS for asset storage - Load balancer Large configuration: - 7 nodes for Control Plane - 3 additional nodes for Monitoring (optional) - NFS for asset storage - Load balancer To ensure high availability and performance, each node must run on its own server in both configurations. If you are using a virtual machine to host the nodes, you must ensure that each VM node is running on a separate physical server. ## Anypoint Platform Core Prerequisites for 4-nodes or 7-nodes to run the Anypoint Platform Core. ### Memory and CPU Requirements The following are the minimum requirements needed for the Anypoint Platform core. Depending on your environment, you may need more CPU or memory.
Table 1. Memory and CPU Requirements
ComponentRequirement

RAM

32 GB

CPU

8 Cores

> [!IMPORTANT] If the stated requirements for Memory and CPU are not met, Anypoint Platform PCE > installation will fail. For production environments, each node in your configuration must have the following dedicated devices. All devices must be formatted either as **xfs** or **ext4**.
Table 2. Dedicated Device Requirements
ComponentSizeMinimum IOPSMount PointDescription

HDD 2 System State Directory Device

250 GB

1500

/var/lib/gravity

Stores system configuration and metadata, including databases and packages. Because package sizes can be large, it is important to estimate the minimum size requirements and allocate enough space as a dedicated device before installation.

HDD 3 Application Data Device

250 GB

1500

/var/lib/data

Stores application configuration and data. The amount of space required should be at least 250 GB, but can vary depending on your specific use case. It is important to estimate the minimum size requirements and allocate enough space as a dedicated device ahead of time.

HDD 4 Docker Device

100 GB

3000

/var/lib/gravity/planet/docker

Used by Docker’s overlay storage driver. A minimum of 100GB is required for the overlay directory. Devices with 50GB or less will experience degraded system performance or might not work at all.

HDD 5 Etcd Device

20 GB

3000

/var/lib/gravity/planet/etcd

Provides dedicated storage for the distributed database used for cluster coordination.

/tmp (Installer)

50 GB

N/A

/ (root)

50 GB

N/A

You must have at least 50GB in your home directory to download and unzip the installer file.

> [!NOTE] The space requirements listed in the previous table are based on average use. For > environments with heavy traffic, validate the amount of space needed with MuleSoft Professional > Services. ### System State Directory Device The main purpose of the system state directory device is to store system configuration and metadata, for example, database and packages. As package sizes can be arbitrarily large, it is important to estimate the minimum size requirements and allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mounted as `/var/lib/gravity`. The following shell snippet is provided as an example for mounting the system state directory device using systemd mount files and cannot be used as-is in production. You can also mount the system state directory device by using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example to guide this process, make sure to specify the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/gravity echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/gravity\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-gravity.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-gravity.mount sudo systemctl start var-lib-gravity.mount ### Application Data Device The main purpose of the application data directory device is to store application configuration and data. The minimum amount of space required is 250GB, but your use case might require more. It is important to allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mounted as `/var/lib/data`. The following shell snippet is provided as an example for mounting the application data device using systemd mount files, and cannot be used as-is in production. You can also mount the application data device using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example to guide this process, make sure to specify the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/data echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/data\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-data.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-data.mount sudo systemctl start var-lib-data.mount ### Etcd Device The main purpose of the etcd device is to provide dedicated storage for a distributed database used for cluster coordination. It does not require much space, 20GB is adequate. Format this device either as `xfs` or `ext4` and mounted as `/var/lib/gravity/planet/etcd`. The following shell snippet mounts the etcd device using systemd mount files to the path `/var/lib/gravity/planet/etcd`. You can also mount the etcd device using `/etc/fstab` or another method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, specify the correct device name in two places. > [!NOTE] Mount this device AFTER you mount the System State device under `/var/lib/gravity/planet`. > If you are using systemd units to mount the volumes, declare mount dependencies using > `After=.`. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/gravity/planet/etcd echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/gravity/planet/etcd\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-gravity-planet-etcd.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-gravity-planet-etcd.mount sudo systemctl start var-lib-gravity-planet-etcd.mount ### Docker Device Unless specified, Docker configuration defaults to the use of the overlay driver, which is recommended for production. Format this device as `xfs` or `ext4` and mount as `/var/lib/gravity/planet/docker`. The following example shell snippet mounts the Docker device using `systemd` mount files to the path `/var/lib/gravity/planet/docker`. You can also mount the Docker device using `/etc/fstab` or another method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. > [!NOTE] Mount this device AFTER you mount the System State device under `/var/lib/gravity/planet`. > If you are using system.d units to mount the volumes, declare mount dependencies using > `After=.` sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/gravity/planet/docker echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/gravity/planet/docker\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-gravity-planet-docker.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-gravity-planet-docker.mount sudo systemctl start var-lib-gravity-planet-docker.mount ## Anypoint Monitoring Anypoint Monitoring is an optional add-on that Anypoint Private Cloud supports. See [Anypoint Monitoring on Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/anypoint-monitoring) for more information. Prerequisites: The extra 3 nodes for the Monitoring Center Add-On. ### Memory and CPU Requirements Monitoring Center Add-On hardware requirements are different from the Platform Core nodes, they require more CPU, more memory and different disk mounts.
Table 3. Memory and CPU Requirements
ComponentRequirement

RAM

128 GB

CPU

32 Cores

For production environments, each node in your configuration must have the following dedicated devices. All devices must be formatted either as **xfs** or **ext4**.
Table 4. Dedicated Device Requirements
ComponentSizeMinimum IOPSMount PointDescription

HDD 2 System State Directory Device

250 GB

1500

/var/lib/gravity

Stores system configuration and metadata, including databases and packages. Package sizes can be large, so it is important to estimate the minimum size requirements and allocate enough space as a dedicated device before installation.

HDD 3 Docker Device

100 GB

3000

/var/lib/gravity/planet/docker

Used by Docker’s overlay storage driver. A minimum of 100GB is required for the overlay directory. Devices with 50GB or less will experience degraded system performance or might not work at all.

HDD 4 Cassandra AMV Device

500 GB

1500

/var/lib/data/dias-meld & /var/lib/data/dias-cassandra

Stores Cassandra database data for Monitoring Center Add-On. The amount of space required should be at least 250 GB, but can vary depending on your specific use case. It is important to estimate the minimum size requirements and allocate enough space as a dedicated device ahead of time.

HDD 5 Stolon AMV Device

250 GB

1500

/var/lib/data/stolon-amv

Stores Postgres database data for Monitorng Center Add-On. The amount of space required should be at least 250 GB, but can vary depending on your specific use case. It is important to estimate the minimum size requirements and allocate enough space as a dedicated device ahead of time.

HDD 7 InfluxDB AMV Device

4 TB

1500

/var/lib/data/influxdb

Stores InfluxDB data for Monitoring Center Add-On. The amount of space required should be at least 250 GB, but can vary depending on your specific use case. It is important to estimate the minimum size requirements and allocate enough space as a dedicated device ahead of time.

HDD 7 logstash AMV Device

500 GB

1500

/var/lib/data/logstash

Stores Logstash data for Monitoring Center Add-On. The amount of space required should be at least 250 GB, but can vary depending on your specific use case. It is important to estimate the minimum size requirements and allocate enough space as a dedicated device ahead of time.

/tmp (Installer)

20 GB

N/A

/ (root)

10 GB

N/A

You must have at least 10GB in your home directory to download and unzip the installer file.

> [!NOTE] The space requirements listed in the previous table are based on average use. For > environments with heavy traffic, validate the amount of space needed with MuleSoft Professional > Services. ### System State Directory Device The main purpose of the system state directory device is to store system configuration and metadata, for example, database and packages. As package sizes can be arbitrarily large, it is important to estimate the minimum size requirements and allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mount as `/var/lib/gravity`. The following shell snippet is provided as an example for mounting the system state directory device using `systemd` mount files and cannot be used as-is in production. You can also mount the system state directory device by using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/gravity echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/gravity\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-gravity.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-gravity.mount sudo systemctl start var-lib-gravity.mount ### Docker Device Unless specified, Docker configuration defaults to the use of the overlay driver, which is recommended for production. Format this device as `xfs` or `ext4` and mount as `/var/lib/gravity/planet/docker`. The following example shell snippet mounts the Docker device using `systemd` mount files to the path: `/var/lib/gravity/planet/docker`. You can also mount the Docker device using `/etc/fstab` or another method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. > [!NOTE] Mount this device AFTER you mount the System State device under `/var/lib/gravity/planet`. > If you are using system.d units to mount the volumes, declare mount dependencies using > `After=.` sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/gravity/planet/docker echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/gravity/planet/docker\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-gravity-planet-docker.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-gravity-planet-docker.mount sudo systemctl start var-lib-gravity-planet-docker.mount ### DIAS Data Device The main purpose of the DIAS Data directory device is to store Monitoring Center Add-On Cassandra and Meld database data. The minimum amount of space required is 250GB, but your use case might require more. It is important to allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mount as `/var/lib/data`. The following shell snippet is provided as an example for mounting the application data device using `systemd` mount files, and cannot be used as-is in production. You can also mount the application data device using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/data echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/data\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-data-dias.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-data-dias.mount sudo systemctl start var-lib-data-dias.mount Once mounted, create the following dirs if necessary on every Monitor Center Add-On node: - /var/lib/data/dias-cassandra - /var/lib/data/dias-meld ### Stolon AMV Device The main purpose of the Stolon AMV directory device is to store Monitoring Center Add-On Postgres database data. The minimum amount of space required is 250GB, but your use case might require more. It is important to allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mount as `/var/lib/data/stolon-amv`. The following shell snippet is provided as an example for mounting the application data device using `systemd` mount files, and cannot be used as-is in production. You can also mount the application data device using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/data/stolon-amv echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/data/stolon-amv\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-data-stolon-amv.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-data-stolon-amv.mount sudo systemctl start var-lib-data-stolon-amv.mount ### InfluxDB AMV Device The main purpose of the InfluxDB AMV directory device is to store Monitoring Center Add-On InfluxDB data. The minimum amount of space required is 250GB, but your use case might require more. It is important to allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mount as `/var/lib/data/influxdb`. The following shell snippet is provided as an example for mounting the application data device using `systemd` mount files, and cannot be used as-is in production. You can also mount the application data device using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/data/influxdb echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/data/influxdb\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-data-influxdb.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-data-influxdb.mount sudo systemctl start var-lib-data-influxdb.mount Once mounted, create the following dirs if necessary on every Monitor Center Add-On node: - /var/lib/data/influxdb/data - /var/lib/data/influxdb/meta ### Logstash AMV Device The main purpose of the Logstash AMV directory device is to store Monitoring Center Add-On Logstash database data. The minimum amount of space required is 250GB, but your use case might require more. It is important to allocate enough space for the dedicated device. Format this device either as `xfs` or `ext4` and mount as `/var/lib/data/logstash`. The following shell snippet is provided as an example for mounting the application data device using `systemd` mount files, and cannot be used as-is in production. You can also mount the application data device using `/etc/fstab` or some other method. Consult with your system administrator to determine the best way to meet this requirement. If you use the following example as a guide for this process, replace `` with the correct device name in two places. sudo mkfs.ext4 /dev/ sudo mkdir -p /var/lib/data/logstash echo -e "\[Mount\]\\nWhat=/dev/\\nWhere=/var/lib/data/logstash\\nType=ext4\\n\[Install\]\\nWantedBy=local-fs.target" | sudo tee /etc/systemd/system/var-lib-data-logstash.mount sudo systemctl daemon-reload sudo systemctl enable var-lib-data-logstash.mount sudo systemctl start var-lib-data-logstash.mount ## Verify Devices To verify that all devices are set up the same way on all nodes, run the `lsblk` command on the Anypoint Monitoring and Visualizer nodes. Output from `lsblk` should resemble the following example: \[root@ip-0-0-0-0 ~\]# lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT xvda 202:0 0 90G 0 disk ├─xvda1 202:1 0 1M 0 part └─xvda2 202:2 0 90G 0 part / xvdb 202:16 0 100G 0 disk /var/lib/gravity/planet/docker xvdd 202:48 0 250G 0 disk /var/lib/data xvde 202:64 0 250G 0 disk /var/lib/gravity xvdf 202:64 0 250G 0 disk /var/lib/data/stolon-amv xvdg 202:64 0 250G 0 disk /var/lib/data/influxdb xvdh 202:64 0 250G 0 disk /var/lib/data/logstash xvdi 202:64 0 20G 0 disk /tmp ## Verify Disk IOPS Depending on your hardware or virtualization provider, you can configure disk IOPS (I/O operations per second). Using the `iops` command, verify available IOPS on each device: $ sudo ./iops --time 2 /dev/xvdb /dev/xvdb, 107.37 GB, 32 threads: 512 B blocks: 1893.0 IO/s, 946.5 KiB/s ( 7.8 Mbit/s) 1 KiB blocks: 1354.8 IO/s, 1.3 MiB/s ( 11.1 Mbit/s) 2 KiB blocks: 1091.8 IO/s, 2.1 MiB/s ( 17.9 Mbit/s) 4 KiB blocks: 807.1 IO/s, 3.2 MiB/s ( 26.4 Mbit/s) 8 KiB blocks: 803.7 IO/s, 6.3 MiB/s ( 52.7 Mbit/s) 16 KiB blocks: 787.4 IO/s, 12.3 MiB/s (103.2 Mbit/s) 32 KiB blocks: 700.8 IO/s, 21.9 MiB/s (183.7 Mbit/s) 64 KiB blocks: 590.0 IO/s, 36.9 MiB/s (309.3 Mbit/s) 128 KiB blocks: 327.6 IO/s, 40.9 MiB/s (343.5 Mbit/s) ... ## Load Balancer Requirements Anypoint Platform PCE must run in a production environment with multiple servers. To distribute traffic among servers and to restrict access to specific ports, you must provide and configure a load balancer. You can use any load balancer, including NGINX, but NGINX is not required. > [!NOTE] If using the [AWS Provisioner for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/prereq-aws-terraform), you do not > need to configure a separate load balancer. An Amazon ELB with the required configuration is > created during installation. ### Load Balancer Configuration You can configure your load balancer to use any method for distributing client requests, but in most contexts a round robin strategy is ideal. Make sure the load balancer is accessible through an IP address that is available to all machines on your network (including the Anypoint Platform PCE nodes) Your load balancer must route the following TCP ports:
Table 5. TCP Port Routing Requirements
Load Balancer PortInstance PortInternal Usage

80

30080

HTTP redirects to HTTPs

443

30443

HTTPS port

8083

30883

HTTPS port for Mule runtimes to authenticate and renew certificates

8889

30889

WebSocket port for Mule runtimes to connect

8895

30895

(Only when Monitoring Center is enabled)

9500

32009

Ops Center access port

In each case, your load balancer must listen on the load balancer port and redirect incoming requests to the instance port. Your Anypoint Platform PCE installation includes an internal NGINX server that listens on each of the configured instance ports, and then performs the action listed in the Internal Usage column. Your load balancer should poll the address `HTTP:10256/healthz` to run a health check on your platform servers and confirm that they are accessible. > [!NOTE] Do not configure SSL certificates in your load balancer. TLS termination is handled by the > platform with the certificates configured using Access Management. See > [Configure Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/config-workflow). ### Implement a Load Balancer Using NGINX 1. Enable stream block in your `/etc/nginx/nginx.conf` file by referencing all the configuration properties defined in `/etc/nginx/stream.d/*`. ```json stream { include /etc/nginx/stream.d/*.conf; } ``` 2. Delete the `default.conf` file from `/etc/nginx/conf.d` 3. Create a folder named `/etc/nginx/stream.d` and in it create a file named `onprem.conf` 4. Paste the following content in `onprem.conf` ```json server { listen 80; proxy_pass 0.0.0.0:30080; } server { listen 443; proxy_pass 0.0.0.0:30443; } server { listen 8083; proxy_pass 0.0.0.0:30883; } server { listen 8889; proxy_pass 0.0.0.0:30889; } server { listen 8895; proxy_pass 0.0.0.0:30895; } server { listen 9500; proxy_pass 0.0.0.0:32009; } server { listen 9501; proxy_pass 0.0.0.0:30083; } ``` 5. Enable the default ports on SELinux before starting the NGINX. Run the following commands one by one: semanage port -a -t http\_port\_t -p tcp 8083 semanage port -a -t http\_port\_t -p tcp 8889 semanage port -a -t http\_port\_t -p tcp 8895 semanage port -a -t http\_port\_t -p tcp 9500 semanage port -a -t http\_port\_t -p tcp 9501 6. Start NGINX: service nginx restart ## See Also - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-platform) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) - [Nginx](https://www.f5.com/products/nginx) - [Anypoint Monitoring](https://docs.mulesoft.com/private-cloud/3.2/anypoint-monitoring) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Installation Workflow > For the full documentation index, see: https://docs.mulesoft.com/llms.txt ## Install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) To install Anypoint Platform PCE, MuleSoft professional services must be involved throughout the installation process. Prior to installation, have your infrastructure team verify that the required infrastructure is correctly installed and configured as described in [Product Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/install-checklist). ## Install Anypoint Monitoring and Anypoint Visualizer After installing Anypoint Platform PCE, contact MuleSoft Professional Services to enable Anypoint Monitoring and Anypoint Visualizer. After they are enabled, you must install the agent for each of these product components. See [Monitor Hybrid Runtimes](https://docs.mulesoft.com/monitoring/am-installing) and [Set up Anypoint Visualizer](https://docs.mulesoft.com/visualizer/setup). > [!NOTE] Multi-instance operation of On-Premises Anypoint Runtime Manager is available only if you > purchase Anypoint Platform Base Subscription, On-premises Management, API Manager and API > Analytics, and Platinum Subscription all together. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Installing Anypoint Platform PCE on OpenShift > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) on OpenShift configure Security Context Constraints (SCC) to grant the required permissions for pods and create roles for Prometheus, allowing it to collect metrics from the services. After these configurations are complete, proceed with installing Anypoint Platform PCE. > [!NOTE] You must have an operational OpenShift cluster prior to installing Anypoint Platform PCE. > If you do not have one set up, refer to the OpenShift documentation for setup instructions. Contact MuleSoft Professional Services to perform the installation of Anypoint Platform PCE. ## Configure the Openshift Environment To configure the OpenShift Environment, configure Security Context Constraints (SCC) and roles for Prometheus. ### Configure Security Context Constraints (SCC) The following snippet contains the SCC configuration: ```yaml apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: true allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: RunAsAny seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc namespace: default rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc subjects: - kind: ServiceAccount name: anypoint-install namespace: default - kind: ServiceAccount name: anypoint-install namespace: dias - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: default namespace: access-management - kind: ServiceAccount name: authentication-server-service-account namespace: access-management - kind: ServiceAccount name: default namespace: design-center - kind: ServiceAccount name: default namespace: api-console-proxy - kind: ServiceAccount name: default namespace: amf - kind: ServiceAccount name: default namespace: arm - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: mozart - kind: ServiceAccount name: default namespace: mocking - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: audit-log - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: api-designer - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: xapi-service-deployment-manager namespace: exchange - kind: ServiceAccount name: default namespace: core-paas - kind: ServiceAccount name: default namespace: monitoring-center - kind: ServiceAccount name: default namespace: visualizer - kind: ServiceAccount name: secrets-robot namespace: dias - kind: ServiceAccount name: default namespace: dias - kind: ServiceAccount name: default namespace: amc - kind: Group name: system:serviceaccounts:monitoring - kind: ServiceAccount name: default namespace: anypoint-flash ``` Save this file as `openshift-scc.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-scc.yaml ``` ### Configure Roles for Prometheus Prometheus requires specific roles to collect metrics from Anypoint Platform PCE services. The following snippet provides the necessary roles configuration: ```yaml apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: pce-prometheus rules: - apiGroups: - "" resources: - endpoints - pods - services - nodes - secrets verbs: - get - list - watch --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: pce-prometheus roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: pce-prometheus subjects: - kind: ServiceAccount name: prometheus-k8s namespace: openshift-monitoring ``` Save this file as `openshift-monitoring.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-monitoring.yaml ``` ## Install Anypoint Platform PCE on OpenShift The following snippet provides an example installation manifest for deploying Anypoint Platform PCE on an OpenShift cluster: ```yaml containerRegistry: endpoint: registry.pce.example.com username: username password: password certificate: "" subproject: example tlsSkipVerify: true postInstallConfiguration: platformDNS: anypoint.pce.example.com fileSystemDNS: fs.pce.example.com firstUserAccount: organizationName: Example Org username: username email: username@mulesoft.com password: password platformCertificate: cert.pem platformCertificateKey: key.pem amvInstallConfiguration: enabled: false storageConfiguration: pceobjectstore: type: persistentVolumeClaim storageClass: example-csi stolon: type: persistentVolumeClaim storageClass: example-csi monitoringAppConfiguration: monitoringStackEnabled: false grafana: enabled: true values: | datasources: datasources.yaml: apiVersion: 1 datasources: - name: Prometheus type: prometheus url: https://prometheus-k8s.openshift-monitoring.svc:9091 access: proxy isDefault: true jsonData: tlsSkipVerify: true httpHeaderName1: 'Authorization' secureJsonData: httpHeaderValue1: 'Bearer ' ``` Note the `monitoringAppConfiguration` section in the snippet. This section disables the monitoring stack for Anypoint Platform PCE and enables only the Grafana application. It is recommended to disable the monitoring stack because OpenShift includes Prometheus by default. Enable Grafana and configure its datasource to connect to Prometheus. To obtain the token for connecting to Prometheus, run this command: ```none oc create token prometheus-k8s -n openshift-monitoring --duration=87600h ``` After completing the installation, run the following commands to allow Prometheus in OpenShift to collect metrics from platform services: ```none oc label namespace default openshift.io/cluster-monitoring="true" --overwrite oc label namespace pce openshift.io/cluster-monitoring="true" --overwrite ``` ## See Also - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Installing Anypoint Platform PCE on OpenShift > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) on OpenShift configure Security Context Constraints (SCC) to grant the required permissions for pods and create roles for Prometheus, allowing it to collect metrics from the services. After these configurations are complete, proceed with installing Anypoint Platform PCE. > [!NOTE] You must have an operational OpenShift cluster prior to installing Anypoint Platform PCE. > If you do not have one set up, refer to the OpenShift documentation for setup instructions. Contact MuleSoft Professional Services to perform the installation of Anypoint Platform PCE. > [!NOTE] Anypoint Platform PCE 4.1.0 is installed and validated on OpenShift 4.16. ## Configure the Openshift Environment To configure the OpenShift Environment, configure Security Context Constraints (SCC) and roles for Prometheus. ### Configure Security Context Constraints (SCC) This snippet contains the SCC configuration: ```yaml apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: false allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: MustRunAs uid: 2020 seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc namespace: pce-core rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc subjects: - kind: ServiceAccount name: pce-cluster-ssl-app namespace: pce-core - kind: ServiceAccount name: default namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: dias - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: default namespace: access-management - kind: ServiceAccount name: authentication-server-service-account namespace: access-management - kind: ServiceAccount name: default namespace: api-console-proxy - kind: ServiceAccount name: default namespace: amf - kind: ServiceAccount name: default namespace: arm - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: mozart - kind: ServiceAccount name: default namespace: mocking - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: audit-log - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: api-designer - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: xapi-service-deployment-manager namespace: exchange - kind: ServiceAccount name: default namespace: core-paas - kind: ServiceAccount name: default namespace: monitoring-center - kind: ServiceAccount name: default namespace: visualizer - kind: ServiceAccount name: secrets-robot namespace: dias - kind: ServiceAccount name: default namespace: dias - kind: ServiceAccount name: default namespace: amc - kind: Group name: system:serviceaccounts:monitoring - kind: ServiceAccount name: default namespace: anypoint-flash --- apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid-privileged allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: true allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: RunAsAny seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolumeClaim - projected - secret - nfs --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc-privileged namespace: pce-core rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid-privileged resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc-privileged roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc-privileged subjects: - kind: ServiceAccount name: stolon namespace: pce-core - kind: ServiceAccount name: default namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: pce-core - kind: ServiceAccount name: seaweedfs-rw-sa namespace: pce-core - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: design-center - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: default namespace: logging-internal - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: mocking --- apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-pce-monitoring allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: false allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: MustRunAs uid: 472 seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-pce-monitoring rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-pce-monitoring resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-pce-monitoring roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-pce-monitoring subjects: - kind: ServiceAccount name: grafana namespace: monitoring ``` Save this file as `openshift-scc.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-scc.yaml ``` ### Configure Roles for Prometheus Prometheus requires specific roles to collect metrics from Anypoint Platform PCE services. The following snippet provides the necessary roles configuration: ```yaml apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: pce-prometheus rules: - apiGroups: - "" resources: - endpoints - pods - services - nodes - secrets verbs: - get - list - watch --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: pce-prometheus roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: pce-prometheus subjects: - kind: ServiceAccount name: prometheus-k8s namespace: openshift-monitoring ``` Save this file as `openshift-monitoring.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-monitoring.yaml ``` ## Install Anypoint Platform PCE on OpenShift The following snippet provides an example installation manifest for deploying Anypoint Platform PCE on an OpenShift cluster: ```yaml containerRegistry: endpoint: registry.pce.example.com username: username password: password certificate: "" subproject: example tlsSkipVerify: true postInstallConfiguration: platformDNS: anypoint.pce.example.com fileSystemDNS: fs.pce.example.com firstUserAccount: organizationName: Example Org username: username email: username@mulesoft.com password: password platformCertificate: cert.pem platformCertificateKey: key.pem amvInstallConfiguration: enabled: false storageConfiguration: pceobjectstore: type: persistentVolumeClaim storageClass: example-csi stolon: type: persistentVolumeClaim storageClass: example-csi monitoringAppConfiguration: monitoringStackEnabled: false grafana: enabled: true values: | datasources: datasources.yaml: apiVersion: 1 datasources: - name: Prometheus type: prometheus url: https://prometheus-k8s.openshift-monitoring.svc:9091 access: proxy isDefault: true jsonData: tlsSkipVerify: true httpHeaderName1: 'Authorization' secureJsonData: httpHeaderValue1: 'Bearer ' ``` Note the `monitoringAppConfiguration` section in the snippet. This section disables the monitoring stack for Anypoint Platform PCE and enables only the Grafana application. It is recommended to disable the monitoring stack because OpenShift includes Prometheus by default. Enable Grafana and configure its datasource to connect to Prometheus. To obtain the token for connecting to Prometheus, run this command: ```none oc create token prometheus-k8s -n openshift-monitoring --duration=87600h ``` After completing the installation, run the following commands to allow Prometheus in OpenShift to collect metrics from platform services: ```none oc label namespace default openshift.io/cluster-monitoring="true" --overwrite oc label namespace pce openshift.io/cluster-monitoring="true" --overwrite ``` ## See Also - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Installing Anypoint Platform PCE on OpenShift > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) on OpenShift configure Security Context Constraints (SCC) to grant the required permissions for pods and create roles for Prometheus, allowing it to collect metrics from the services. After these configurations are complete, proceed with installing Anypoint Platform PCE. > [!NOTE] Contact MuleSoft Professional Services or your Anypoint PCE Certified Partner to perform > the installation or upgrade of Anypoint Platform PCE. > [!NOTE] You must have an operational OpenShift cluster prior to installing Anypoint Platform PCE. > If you do not have one set up, refer to the OpenShift documentation for setup instructions. > [!NOTE] Anypoint Platform PCE 4.2.1 is installed and validated on OpenShift 4.18 and 4.20. > [!NOTE] It isn’t supported to in-place upgrade from an existing Openshift PCE 4.1.x cluster that’s > been installed with full admin privilege to 4.2.1 or higher with no admin scope. If this is your > case, consult [Migrating Anypoint Platform PCE to v4.x](https://docs.mulesoft.com/private-cloud/4.2/upgrade). ## Configure the Openshift Environment To configure the OpenShift Environment, configure Security Context Constraints (SCC) and roles for Prometheus. ### Configure Security Context Constraints (SCC) This snippet contains the SCC configuration: ```yaml apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: false allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: MustRunAs uid: 2020 seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolume - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc namespace: pce-core rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc subjects: - kind: ServiceAccount name: pce-cluster-ssl-app namespace: pce-core - kind: ServiceAccount name: default namespace: pce-core - kind: ServiceAccount name: anypoint namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: pce-core - kind: ServiceAccount name: anypoint-install-anypoint-services namespace: pce-core - kind: ServiceAccount name: anypoint-install-cluster-ssl namespace: pce-core - kind: ServiceAccount name: anypoint-install-namespaces-job namespace: pce-core - kind: ServiceAccount name: anypoint-install-seaweedfs namespace: pce-core - kind: ServiceAccount name: anypoint-prometheus-install namespace: pce-core - kind: ServiceAccount name: anypoint-grafana-install namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: dias - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: default namespace: access-management - kind: ServiceAccount name: authentication-server-service-account namespace: access-management - kind: ServiceAccount name: default namespace: api-console-proxy - kind: ServiceAccount name: default namespace: amf - kind: ServiceAccount name: default namespace: arm - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: mozart - kind: ServiceAccount name: default namespace: mocking - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: audit-log - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: api-designer - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: xapi-service-deployment-manager namespace: exchange - kind: ServiceAccount name: default namespace: core-paas - kind: ServiceAccount name: default namespace: monitoring-center - kind: ServiceAccount name: default namespace: visualizer - kind: ServiceAccount name: secrets-robot namespace: dias - kind: ServiceAccount name: default namespace: dias - kind: ServiceAccount name: default namespace: amc - kind: Group name: system:serviceaccounts:monitoring - kind: ServiceAccount name: default namespace: anypoint-flash - kind: ServiceAccount name: default namespace: design-center --- apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid-privileged allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: true allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: RunAsAny seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolume - persistentVolumeClaim - projected - secret - nfs --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc-privileged rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid-privileged resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc-privileged roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc-privileged subjects: - kind: ServiceAccount name: stolon namespace: pce-core - kind: ServiceAccount name: default namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: pce-core - kind: ServiceAccount name: seaweedfs-rw-sa namespace: pce-core - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: design-center - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: default namespace: logging-internal - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: mocking --- apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-pce-monitoring allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: false allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: MustRunAs uid: 472 seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-pce-monitoring rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-pce-monitoring resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-pce-monitoring roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-pce-monitoring subjects: - kind: ServiceAccount name: grafana namespace: monitoring ``` Save this file as `openshift-scc.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-scc.yaml ``` ### Configure Roles for Prometheus Prometheus requires specific roles to collect metrics from Anypoint Platform PCE services. The following snippet provides the necessary roles configuration: ```yaml apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: pce-prometheus rules: - apiGroups: - "" resources: - endpoints - pods - services - nodes - secrets verbs: - get - list - watch --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: pce-prometheus roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: pce-prometheus subjects: - kind: ServiceAccount name: prometheus-k8s namespace: openshift-monitoring ``` Save this file as `openshift-monitoring.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-monitoring.yaml ``` ## Install Anypoint Platform PCE on OpenShift Create an `input.yaml` file using this template and update it with your environment details. This template matches the PCE 4.2 installation configuration, with monitoring settings for OpenShift: ```yaml # Container Registry configuration to upload PCE images containerRegistry: endpoint: "" # Container Registry Endpoint username: "" # Provide either Username/Password OR Certificate to authenticate with container registry password: "" certificate: "/etc/registry/certificate/ca.crt" subproject: "mulesoft" # Container registry subproject under which all the PCE images will be part of. tlsSkipVerify: True # True or False. # Platform configuration using CSI driver for filesystem access platformConfiguration: platformDNS: "" fileSystemCsiDriverName: "" # CSI driver name (e.g., nfs.csi.k8s.io, efs.csi.aws.com) # Provide NFS DNS and path (not required if using EFS CSI Driver) # fileSystemDNS: "" # fileSystemPath: "" # Provide individual NFS volume handles (all 4 are required if using EFS CSI Driver. You'll need to create 4 access points pointing to the same dir in the EFS) fileSystemWcCsiVolumeHandle: "" # CSI volume handle for wc filesystem. E.g.: fs-0835b0d1eb7588eb1::fsap-03f6c68563ceb5acf fileSystem01CsiVolumeHandle: "" # CSI volume handle for 01 filesystem fileSystemBareCsiVolumeHandle: "" # CSI volume handle for bare filesystem fileSystemBackupRestoreCsiVolumeHandle: "" # CSI volume handle for backup-restore filesystem platformCertificate: /path/to/platform/certificate/cert.crt platformCertificateKey: path/to/platform/certificate/cert.key firstUserAccount: organizationName: "Test Org" username: "username" email: "username@mulesoft.com" password: "" # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "10Gi" # Optional. Defaults to 10Gi volumeSize: "100Gi" # Optional. Defaults to 100Gi volumeIndexSize: "10Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "100Gi" # Optional. Defaults to 100Gi backupRestoreTmp: storageClass: "" # Storage class for backup/restore temporary storage # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStack: enabled: False # Defaults to True. Must be set to False for Openshift clusters. # Grafana configuration for OpenShift. Disable the monitoring stack because OpenShift includes Prometheus by default. grafana: enabled: True # Defaults to False. values: | datasources: datasources.yaml: apiVersion: 1 datasources: - name: Prometheus type: prometheus url: https://prometheus-k8s.openshift-monitoring.svc:9091 access: proxy isDefault: true jsonData: tlsSkipVerify: true httpHeaderName1: 'Authorization' secureJsonData: httpHeaderValue1: 'Bearer ' ``` In the `monitoringAppConfiguration` section, set `monitoringStack.enabled` to `False` and enable only Grafana. Disable the Anypoint Platform PCE monitoring stack because OpenShift includes Prometheus by default. Enable Grafana and configure its datasource to connect to Prometheus. To obtain the token for connecting to Prometheus, run this command: ```none oc create token prometheus-k8s -n openshift-monitoring --duration=87600h ``` After you complete the installation, run these commands to allow Prometheus in OpenShift to collect metrics from platform services: ```none oc label namespace default openshift.io/cluster-monitoring="true" --overwrite oc label namespace pce openshift.io/cluster-monitoring="true" --overwrite ``` ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.2/install-checklist) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Installing Anypoint Platform PCE on OpenShift > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) on OpenShift configure Security Context Constraints (SCC) to grant the required permissions for pods and create roles for Prometheus, allowing it to collect metrics from the services. After these configurations are complete, proceed with installing Anypoint Platform PCE. > [!NOTE] Contact MuleSoft Professional Services or your Anypoint PCE Certified Partner to perform > the installation or upgrade of Anypoint Platform PCE. > [!NOTE] You must have an operational OpenShift cluster prior to installing Anypoint Platform PCE. > If you do not have one set up, refer to the OpenShift documentation for setup instructions. > [!NOTE] Anypoint Platform PCE 4.2.1 is installed and validated on OpenShift 4.18 and 4.20. > [!NOTE] It isn’t supported to in-place upgrade from an existing Openshift PCE 4.1.x cluster that’s > been installed with full admin privilege to 4.2.1 or higher with no admin scope. If this is your > case, consult [Migrating Anypoint Platform PCE to v4.x](https://docs.mulesoft.com/private-cloud/latest/upgrade). ## Configure the Openshift Environment To configure the OpenShift Environment, configure Security Context Constraints (SCC) and roles for Prometheus. ### Configure Security Context Constraints (SCC) This snippet contains the SCC configuration: ```yaml apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: false allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: MustRunAs uid: 2020 seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolume - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc namespace: pce-core rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc subjects: - kind: ServiceAccount name: pce-cluster-ssl-app namespace: pce-core - kind: ServiceAccount name: default namespace: pce-core - kind: ServiceAccount name: anypoint namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: pce-core - kind: ServiceAccount name: anypoint-install-anypoint-services namespace: pce-core - kind: ServiceAccount name: anypoint-install-cluster-ssl namespace: pce-core - kind: ServiceAccount name: anypoint-install-namespaces-job namespace: pce-core - kind: ServiceAccount name: anypoint-install-seaweedfs namespace: pce-core - kind: ServiceAccount name: anypoint-prometheus-install namespace: pce-core - kind: ServiceAccount name: anypoint-grafana-install namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: dias - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: default namespace: access-management - kind: ServiceAccount name: authentication-server-service-account namespace: access-management - kind: ServiceAccount name: default namespace: api-console-proxy - kind: ServiceAccount name: default namespace: amf - kind: ServiceAccount name: default namespace: arm - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: mozart - kind: ServiceAccount name: default namespace: mocking - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: audit-log - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: api-designer - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: xapi-service-deployment-manager namespace: exchange - kind: ServiceAccount name: default namespace: core-paas - kind: ServiceAccount name: default namespace: monitoring-center - kind: ServiceAccount name: default namespace: visualizer - kind: ServiceAccount name: secrets-robot namespace: dias - kind: ServiceAccount name: default namespace: dias - kind: ServiceAccount name: default namespace: amc - kind: Group name: system:serviceaccounts:monitoring - kind: ServiceAccount name: default namespace: anypoint-flash - kind: ServiceAccount name: default namespace: design-center --- apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-anyuid-privileged allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: true allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: RunAsAny seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolume - persistentVolumeClaim - projected - secret - nfs --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-scc-privileged rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-anyuid-privileged resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-scc-privileged roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-scc-privileged subjects: - kind: ServiceAccount name: stolon namespace: pce-core - kind: ServiceAccount name: default namespace: pce-core - kind: ServiceAccount name: anypoint-install namespace: pce-core - kind: ServiceAccount name: seaweedfs-rw-sa namespace: pce-core - kind: ServiceAccount name: default namespace: api-manager - kind: ServiceAccount name: default namespace: design-center - kind: ServiceAccount name: default namespace: exchange - kind: ServiceAccount name: default namespace: logging-internal - kind: ServiceAccount name: default namespace: pce - kind: ServiceAccount name: monitoring namespace: pce - kind: ServiceAccount name: default namespace: trusted-domains - kind: ServiceAccount name: default namespace: mocking --- apiVersion: security.openshift.io/v1 kind: SecurityContextConstraints metadata: name: anypoint-pce-monitoring allowHostDirVolumePlugin: false allowHostIPC: false allowHostNetwork: false allowHostPID: false allowHostPorts: false allowPrivilegeEscalation: false allowPrivilegedContainer: false allowedCapabilities: null defaultAddCapabilities: null fsGroup: type: RunAsAny users: [] groups: [] readOnlyRootFilesystem: false requiredDropCapabilities: - MKNOD runAsUser: type: MustRunAs uid: 472 seLinuxContext: type: MustRunAs seccompProfiles: - runtime/default supplementalGroups: type: RunAsAny volumes: - configMap - csi - downwardAPI - emptyDir - ephemeral - persistentVolumeClaim - projected - secret --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: anypoint-pce-monitoring rules: - apiGroups: ["security.openshift.io"] resourceNames: - anypoint-pce-monitoring resources: ["securitycontextconstraints"] verbs: ["use"] --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: anypoint-pce-monitoring roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: anypoint-pce-monitoring subjects: - kind: ServiceAccount name: grafana namespace: monitoring ``` Save this file as `openshift-scc.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-scc.yaml ``` ### Configure Roles for Prometheus Prometheus requires specific roles to collect metrics from Anypoint Platform PCE services. The following snippet provides the necessary roles configuration: ```yaml apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRole metadata: name: pce-prometheus rules: - apiGroups: - "" resources: - endpoints - pods - services - nodes - secrets verbs: - get - list - watch --- apiVersion: rbac.authorization.k8s.io/v1 kind: ClusterRoleBinding metadata: name: pce-prometheus roleRef: apiGroup: rbac.authorization.k8s.io kind: ClusterRole name: pce-prometheus subjects: - kind: ServiceAccount name: prometheus-k8s namespace: openshift-monitoring ``` Save this file as `openshift-monitoring.yaml` and apply it to the OpenShift cluster: ```none oc apply -f openshift-monitoring.yaml ``` ## Install Anypoint Platform PCE on OpenShift The following snippet provides an example installation manifest for deploying Anypoint Platform PCE on an OpenShift cluster: ```yaml containerRegistry: endpoint: registry.pce.example.com username: username password: password certificate: "" subproject: example tlsSkipVerify: true postInstallConfiguration: platformDNS: anypoint.pce.example.com fileSystemDNS: fs.pce.example.com firstUserAccount: organizationName: Example Org username: username email: username@mulesoft.com password: password platformCertificate: cert.pem platformCertificateKey: key.pem amvInstallConfiguration: enabled: false storageConfiguration: pceobjectstore: type: persistentVolumeClaim storageClass: example-csi stolon: type: persistentVolumeClaim storageClass: example-csi monitoringAppConfiguration: monitoringStackEnabled: false grafana: enabled: true values: | datasources: datasources.yaml: apiVersion: 1 datasources: - name: Prometheus type: prometheus url: https://prometheus-k8s.openshift-monitoring.svc:9091 access: proxy isDefault: true jsonData: tlsSkipVerify: true httpHeaderName1: 'Authorization' secureJsonData: httpHeaderValue1: 'Bearer ' ``` Note the `monitoringAppConfiguration` section in the snippet. This section disables the monitoring stack for Anypoint Platform PCE and enables only the Grafana application. It is recommended to disable the monitoring stack because OpenShift includes Prometheus by default. Enable Grafana and configure its datasource to connect to Prometheus. To obtain the token for connecting to Prometheus, run this command: ```none oc create token prometheus-k8s -n openshift-monitoring --duration=87600h ``` After completing the installation, run the following commands to allow Prometheus in OpenShift to collect metrics from platform services: ```none oc label namespace default openshift.io/cluster-monitoring="true" --overwrite oc label namespace pce openshift.io/cluster-monitoring="true" --overwrite ``` ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/latest/install-workflow) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/latest/install-checklist) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Installing Anypoint Platform PCE with Bring Your Own Namespaces > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Starting with Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) 4.3.0, you can install the platform in Bring Your Own Namespaces (BYONS) mode. In a standard installation, the installer creates and names the Kubernetes namespaces it needs. In BYONS mode, your cluster administrator creates and names those namespaces before installation, and Anypoint Platform PCE installs into the namespaces you prepared. Use BYONS when your organization must apply its own namespace names, or when cluster administrators must create and review namespaces before any application is installed into them. Names must be valid Kubernetes namespace names. They can follow your organization’s convention, such as names tied to a cost center or business unit. In both modes, the installed platform is the same. BYONS changes who creates the namespaces and what they are named. ## Considerations Before you install in BYONS mode, consider the following: - BYONS mode is available starting with Anypoint Platform PCE 4.3.0. It is not available in earlier versions. - Namespace names must be valid Kubernetes namespace names. Otherwise, they can follow your organization’s naming convention. - The installer doesn’t create or delete namespaces in BYONS mode. Your organization keeps control of the namespace lifecycle. - The installer validates the cluster before installation starts. Most configuration mistakes are caught before any change is made, and the installer reports what to correct. ## Before You Begin To install in BYONS mode, your team provides the following: - A namespace for every Anypoint Platform PCE component, created before installation. - A label on each namespace so Anypoint Platform PCE can identify its purpose. - A mapping that links each platform component to the namespace name you chose. - Cluster-level permissions applied by your cluster administrator before installation begins. The installer doesn’t have broad cluster access in BYONS mode. Additionally, your MuleSoft representative provides the BYONS installation package. The package includes the list of required namespaces, the required labels, a mapping template. ## Compare BYONS with a Standard Installation
Standard installationBYONS installation

Who creates the namespaces

The Anypoint Platform PCE installer

Your cluster administrator

Who names the namespaces

Anypoint Platform PCE, using fixed names

You, using your own names

Installer cluster access

Cluster-admin

Limited privileges. The installer doesn’t have cluster-admin rights.

When cluster-level resources are set up

During installation

Before installation, by your cluster administrator

## Install in BYONS Mode A BYONS installation is a hand-off between your cluster administrator and the Anypoint Platform PCE installer. Before you start, contact your MuleSoft representative to review your naming conventions and obtain the BYONS installation package. The package includes the list of required namespaces, the required labels, a mapping template. The installation process is as follows: 1. Your cluster administrator creates every namespace Anypoint Platform PCE requires, using your organization’s names, and applies the required labels so the installer can recognize each namespace. Your cluster administrator also applies the cluster-level permissions the platform needs. 2. Your cluster administrator provides a namespace mapping. The mapping tells Anypoint Platform PCE which of your namespace names corresponds to each platform component. 3. The installer validates the cluster before it makes changes. It checks that every required namespace exists, that the names in the mapping match those namespaces, and that each namespace has the expected labels. If something is missing or inconsistent, the installer stops and reports what to correct. 4. After validation succeeds, installation continues through the same phases as a standard installation. Anypoint Platform PCE places its components in the namespaces you created. ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/latest/install-workflow) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/latest/install-checklist) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) > For the full documentation index, see: https://docs.mulesoft.com/llms.txt The process to install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), consist of the following steps: 1. Setting up and verifying your infrastructure 2. Performing a pre-installation verification 3. Installing Anypoint Platform PCE > [!NOTE] Contact MuleSoft Professional Services before performing the pre-installation verification > and installation of Anypoint Platform PCE because they are involved in these processes. > [!NOTE] To install Anypoint Platform PCE on OpenShift, see > [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/4.0/install-openshift). ## Set Up and Verify your Infrastructure Before you install Anypoint Platform PCE, ensure your infrastructure team installed and configured correctly the infrastructure for your organization following the requirements as described in [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist). ## Perform a Pre-installation Verification To perform the pre-installation healthcheck verification for the Kubernetes (K8s) cluster, follow these steps: 1. Confirm your Kubernetes cluster version is 1.30.x or 1.29.x. 2. Create an `input.yaml` file using the following template and update it with your environment details: ```json # Container Registry configuration to upload PCE images containerRegistry: endpoint: "" # Container Registry Endpoint username: "" # Provide either Username/Password OR Certificate to authenticate with container registry password: "" certificate: "/etc/registry/certificate/ca.crt" subproject: "mulesoft" # Container registry subproject under which all the PCE images will be part of. tlsSkipVerify: True # True or False. # Configuration required to perform post installation setup for Anypoint platform # All the fields in `postInstallConfiguration` are required. postInstallConfiguration: platformDNS: "" fileSystemDNS: "" firstUserAccount: organizationName: 'Test Org' username: 'username' email: 'username@mulesoft.com' password: '' platformCertificate: /path/to/platform/certificate/cert.crt platformCertificateKey: path/to/platform/certificate/cert.key # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "10Gi" # Optional. Defaults to 10Gi volumeSize: "100Gi" # Optional. Defaults to 100Gi volumeIndexSize: "10Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "100Gi" # Optional. Defaults to 100Gi # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStackEnabled: True # Defaults to True. Must be set to False for Openshift clusters. grafana: enabled: False # Defaults to False. Set to True for OpenShift clusters. ``` 3. Obtain the `license.lic` file from the MuleSoft Professional Services team. 4. Set the `KUBECONFIG` environment variable to the path to the `kubeconfig` file of the cluster. 5. Verify that the cluster satisfies the prerequisites by running the following command against your `input.yaml` file and the `license.lic` file: `./pcectl pre-install-check --input-file/path/to/input.yaml --license-file /path/to/license.lic` 6. For Anypoint Platform PCE workload to be scheduled on designated nodes, apply the following labels to all worker nodes: - `mulesoft-role=node` - `redis-role=node` 7. Apply the following labels to any three worker nodes. These do not need to be the same three nodes: - `stolon-keeper=yes` - `pce-data-nodes=yes` 8. (Optional) Apply taints to controlplane nodes to ensure that no Anypoint Platform PCE workload is scheduled on the control plane. This also affects any other workload your organization may want to run on control plane nodes. Consult with your Kubernetes cluster admin to perform this step: `mulesoft.pce/controlplane-node=true:NoSchedule` ## Install Anypoint Platform PCE After your infrastructure is correctly set up and you performed a pre-installation verification, MuleSoft Professional Services will assist you with the installation of Anypoint Platform PCE. To install Anypoint Platform PCE, run this command: `./pcectl install --input-file /path/to/input.yaml --license-file /path/to/license.lic` ## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist). --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) > For the full documentation index, see: https://docs.mulesoft.com/llms.txt The process to install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), consist of the following steps: 1. Setting up and verifying your infrastructure 2. Performing a pre-installation verification 3. Installing Anypoint Platform PCE > [!NOTE] Contact MuleSoft Professional Services before performing the pre-installation verification > and installation of Anypoint Platform PCE because they are involved in these processes. > [!NOTE] To install Anypoint Platform PCE on OpenShift, see > [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/4.1/install-openshift). ## Set Up and Verify your Infrastructure Before you install Anypoint Platform PCE, ensure your infrastructure team installed and configured correctly the infrastructure for your organization following the requirements as described in [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist). ## Perform a Pre-installation Verification To perform the pre-installation healthcheck verification for the Kubernetes (K8s) cluster, follow these steps: 1. Confirm your Kubernetes cluster version is 1.32.x or 1.31.x. 2. Create an `input.yaml` file using the following template and update it with your environment details: ```json # Container Registry configuration to upload PCE images containerRegistry: endpoint: "" # Container Registry Endpoint username: "" # Provide either Username/Password OR Certificate to authenticate with container registry password: "" certificate: "/etc/registry/certificate/ca.crt" subproject: "mulesoft" # Container registry subproject under which all the PCE images will be part of. tlsSkipVerify: True # True or False. # All the fields in `platformConfiguration` are required. platformConfiguration: platformDNS: "" fileSystemDNS: "" fileSystemPath: "" platformCertificate: /path/to/platform/certificate/cert.crt platformCertificateKey: path/to/platform/certificate/cert.key firstUserAccount: organizationName: 'Test Org' username: 'username' email: 'username@mulesoft.com' password: '' # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "10Gi" # Optional. Defaults to 10Gi volumeSize: "100Gi" # Optional. Defaults to 100Gi volumeIndexSize: "10Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "100Gi" # Optional. Defaults to 100Gi # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStack: enabled: True # Defaults to True. Must be set to False for Openshift clusters. values: | # Optional. If not provided, the default values will be used. prometheus: prometheusSpec: storageSpec: volumeClaimTemplate: spec: storageClassName: 'local-path', accessModes: ["ReadWriteOnce"] resources: requests: storage: 50Gi ## The following section is used to configure Grafana for an Openshift cluster. It is recommended to disable the monitoring stack as OpenShift includes Prometheus by default. grafana: enabled: False # Defaults to False. values: | datasources: datasources.yaml: apiVersion: 1 datasources: - name: Prometheus type: prometheus url: https://prometheus-k8s.openshift-monitoring.svc:9091 access: proxy isDefault: true jsonData: tlsSkipVerify: true httpHeaderName1: 'Authorization' secureJsonData: httpHeaderValue1: 'Bearer ' ``` 3. Obtain the `license.lic` file from the MuleSoft Professional Services team. 4. Set the `KUBECONFIG` environment variable to the path to the `kubeconfig` file of the cluster. 5. Verify that the cluster satisfies the prerequisites by running the following command against your `input.yaml` file and the `license.lic` file: `./pcectl pre-install-check --input-file/path/to/input.yaml --license-file /path/to/license.lic` 6. For Anypoint Platform PCE workload to be scheduled on designated nodes, apply the following labels to all worker nodes: - `mulesoft-role=node` - `redis-role=node` 7. Apply the following labels to any three worker nodes. These do not need to be the same three nodes: - `stolon-keeper=yes` - `pce-data-nodes=yes` 8. (Optional) Apply taints to controlplane nodes to ensure that no Anypoint Platform PCE workload is scheduled on the control plane. This also affects any other workload your organization may want to run on control plane nodes. Consult with your Kubernetes cluster admin to perform this step: `mulesoft.pce/controlplane-node=true:NoSchedule` ## Install Anypoint Platform PCE After your infrastructure is correctly set up and you performed a pre-installation verification, MuleSoft Professional Services will assist you with the installation of Anypoint Platform PCE. To install Anypoint Platform PCE, run this command: `./pcectl install --input-file /path/to/input.yaml --license-file /path/to/license.lic` ## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist). --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) > For the full documentation index, see: https://docs.mulesoft.com/llms.txt The process to install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), consist of the following steps: 1. Obtaining your Anypoint PCE installation distribution 2. Setting up and verifying your infrastructure 3. Performing a pre-installation verification 4. Installing or Upgrading Anypoint Platform PCE > [!NOTE] Contract Mulesoft Processional Services or your Anypoint PCE Certified Partner to obtain > your Anypoint PCE installation distribution and before performing the pre-installation > verification and installation or upgrade of the Anypoint PCE Platform to ensure they are involved > in these procedures. > [!NOTE] To install Anypoint Platform PCE on OpenShift, see > [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/4.2/install-openshift). ## Set Up and Verify your Infrastructure Before you install Anypoint Platform PCE, ensure your infrastructure team installed and configured correctly the infrastructure for your organization following the requirements as described in [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.2/install-checklist). ## Perform a Pre-installation Verification To perform the pre-installation healthcheck verification for the Kubernetes (K8s) cluster, follow these steps: 1. Confirm your Kubernetes cluster version is 1.34.x or 1.33.x. 2. Create an `input.yaml` file using this template and update it with your environment details: ```json # Container Registry configuration to upload PCE images containerRegistry: endpoint: "" # Container Registry Endpoint username: "" # Provide either Username/Password OR Certificate to authenticate with container registry password: "" certificate: "/etc/registry/certificate/ca.crt" subproject: "mulesoft" # Container registry subproject under which all the PCE images will be part of. tlsSkipVerify: True # True or False. # Platform configuration using CSI driver for filesystem access platformConfiguration: platformDNS: "" fileSystemCsiDriverName: "" # CSI driver name (e.g., nfs.csi.k8s.io, efs.csi.aws.com) # Provide NFS DNS and path (not required if using EFS CSI Driver) # fileSystemDNS: "" # fileSystemPath: "" # Provide individual NFS volume handles (all 4 are required if using EFS CSI Driver. You'll need to create 4 access points pointing to the same dir in the EFS) fileSystemWcCsiVolumeHandle: "" # CSI volume handle for wc filesystem. E.g.: fs-0835b0d1eb7588eb1::fsap-03f6c68563ceb5acf fileSystem01CsiVolumeHandle: "" # CSI volume handle for 01 filesystem fileSystemBareCsiVolumeHandle: "" # CSI volume handle for bare filesystem fileSystemBackupRestoreCsiVolumeHandle: "" # CSI volume handle for backup-restore filesystem platformCertificate: /path/to/platform/certificate/cert.crt platformCertificateKey: path/to/platform/certificate/cert.key firstUserAccount: organizationName: 'Test Org' username: 'username' email: 'username@mulesoft.com' password: '' # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "10Gi" # Optional. Defaults to 10Gi volumeSize: "100Gi" # Optional. Defaults to 100Gi volumeIndexSize: "10Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "100Gi" # Optional. Defaults to 100Gi backupRestoreTmp: storageClass: "" # Storage class for backup/restore temporary storage # Configuration options for monitoring stack monitoringAppConfiguration: monitoringStack: enabled: True # Defaults to True. Must be set to False for Openshift clusters. # values: | # Optional. If not provided, the default values will be used. # prometheus: # prometheusSpec: # storageSpec: # volumeClaimTemplate: # spec: # storageClassName: 'local-path', # accessModes: ["ReadWriteOnce"] # resources: # requests: # storage: 50Gi ## The following section is used to configure Grafana for an Openshift cluster. It is recommended to disable the monitoring stack as OpenShift includes Prometheus by default. # grafana: # enabled: False # Defaults to False. # values: | # datasources: # datasources.yaml: # apiVersion: 1 # datasources: # - name: Prometheus # type: prometheus # url: https://prometheus-k8s.openshift-monitoring.svc:9091 # access: proxy # isDefault: true # jsonData: # tlsSkipVerify: true # httpHeaderName1: 'Authorization' # secureJsonData: # httpHeaderValue1: 'Bearer ' ``` 3. Obtain the `license.lic` file from the MuleSoft Professional Services team. 4. Set the `KUBECONFIG` environment variable to the path to the `kubeconfig` file of the cluster. 5. Verify that the cluster satisfies the prerequisites by running the following command against your `input.yaml` file and the `license.lic` file: `./pcectl pre-install-check --input-file/path/to/input.yaml --license-file /path/to/license.lic` 6. For Anypoint Platform PCE workload to be scheduled on designated nodes, apply the following labels to all worker nodes: - `mulesoft-role=node` - `redis-role=node` 7. Apply the following labels to any three worker nodes. These do not need to be the same three nodes: - `stolon-keeper=yes` - `pce-data-nodes=yes` 8. (Optional) Apply taints to controlplane nodes to ensure that no Anypoint Platform PCE workload is scheduled on the control plane. This also affects any other workload your organization may want to run on control plane nodes. Consult with your Kubernetes cluster admin to perform this step: `mulesoft.pce/controlplane-node=true:NoSchedule` ## Install Anypoint Platform PCE After your infrastructure is correctly set up and you performed a pre-installation verification, MuleSoft Professional Services or your Anypoint PCE Certified Partner will assist you with the installation of Anypoint Platform PCE. To install Anypoint Platform PCE, run this command: `./pcectl install --input-file /path/to/input.yaml --license-file /path/to/license.lic` ## Upgrade Anypoint Platform PCE After you have completed a pre-installation/upgrade verification, MuleSoft Professional Services or your Anypoint PCE Certified Partner will assist you with the upgrade of Anypoint Platform PCE. To upgrade Anypoint Platform PCE, run this command: `./pcectl upgrade /path/to/input.yaml --license-file /path/to/license.lic` ## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.2/install-checklist). --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) > For the full documentation index, see: https://docs.mulesoft.com/llms.txt The process to install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), consist of the following steps: 1. Obtaining your Anypoint PCE installation distribution 2. Setting up and verifying your infrastructure 3. Performing a pre-installation verification 4. Installing or Upgrading Anypoint Platform PCE > [!NOTE] Contract Mulesoft Processional Services or your Anypoint PCE Certified Partner to obtain > your Anypoint PCE installation distribution and before performing the pre-installation > verification and installation or upgrade of the Anypoint PCE Platform to ensure they are involved > in these procedures. > [!NOTE] To install Anypoint Platform PCE on OpenShift, see > [Installing Anypoint Platform PCE on OpenShift](https://docs.mulesoft.com/private-cloud/latest/install-openshift). > [!NOTE] To install Anypoint Platform PCE into namespaces that your organization creates and names, > see [Installing Anypoint Platform PCE with Bring Your Own Namespaces](https://docs.mulesoft.com/private-cloud/latest/install-byons). ## Set Up and Verify your Infrastructure Before you install Anypoint Platform PCE, ensure your infrastructure team installed and configured correctly the infrastructure for your organization following the requirements as described in [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/latest/install-checklist). ## Perform a Pre-installation Verification To perform the pre-installation healthcheck verification for the Kubernetes (K8s) cluster, follow these steps: 1. Confirm your Kubernetes cluster version is 1.35.x, 1.34.x, 1.33.x, or 1.32.x. 2. Create an `input.yaml` file using this template and update it with your environment details: ```json # Container Registry configuration to upload PCE images containerRegistry: endpoint: "" # Container Registry Endpoint username: "" # Provide either Username/Password OR Certificate to authenticate with container registry password: "" certificate: "/etc/registry/certificate/ca.crt" subproject: "mulesoft" # Container registry subproject under which all the PCE images will be part of. tlsSkipVerify: True # True or False. # Platform configuration for the shared ReadWriteMany (RWX) filesystem. # Choose exactly one of the two storage modes below. NFS mode has two configuration # variants (Option A or Option B). StorageClass mode is Option C. The modes are # mutually exclusive. platformConfiguration: platformDNS: "" # ---- NFS mode (Option A and Option B) ---- # Requires a pre-installed CSI driver in the Kubernetes cluster. fileSystemCsiDriverName: "" # CSI driver name (e.g., nfs.csi.k8s.io, efs.csi.aws.com) # Optional NFS mount option overrides. Comma-separated string. Only used in NFS mode. fileSystemBareMountOptionsOverride: "" fileSystemWcMountOptionsOverride: "" fileSystem01MountOptionsOverride: "" # Option A - Generic NFS CSI driver: provide DNS and path. # fileSystemDNS: "" # NFS server hostname or IP # fileSystemPath: "" # NFS export path (e.g., /exports/pce) # Option B - CSI driver with individual volume handles (all 4 required, e.g. EFS CSI Driver # with 4 access points that point to the same directory in the EFS): fileSystemWcCsiVolumeHandle: "" # CSI volume handle for wc filesystem. E.g.: fs-0835b0d1eb7588eb1::fsap-03f6c68563ceb5acf fileSystem01CsiVolumeHandle: "" # CSI volume handle for 01 filesystem fileSystemBareCsiVolumeHandle: "" # CSI volume handle for bare filesystem fileSystemBackupRestoreCsiVolumeHandle: "" # CSI volume handle for backup-restore filesystem # ---- StorageClass mode (Option C, Anypoint Platform PCE 4.3.0 and later) ---- # Mutually exclusive with NFS mode. The StorageClass must support ReadWriteMany. # Leave fileSystemCsiDriverName and all NFS-mode fields above empty when using this option. # For backup retention, use a StorageClass with reclaimPolicy: Retain so deleting the PVC # does not destroy historical backup tarballs. # fileSystemStorageClass: "" # e.g. ocs-storagecluster-cephfs platformCertificate: /path/to/platform/certificate/cert.crt platformCertificateKey: path/to/platform/certificate/cert.key firstUserAccount: organizationName: 'Test Org' username: 'username' email: 'username@mulesoft.com' password: '' # Storage class configuration for persistent data services storageConfiguration: pceobjectstore: type: "persistentVolumeClaim" storageClass: "" controlPlaneSize: "10Gi" # Optional. Defaults to 10Gi volumeSize: "100Gi" # Optional. Defaults to 100Gi volumeIndexSize: "10Gi" # Optional. Defaults to 10Gi stolon: type: "persistentVolumeClaim" storageClass: "" size: "100Gi" # Optional. Defaults to 100Gi backupRestoreTmp: storageClass: "" # Storage class for backup/restore temporary storage ``` 3. Obtain the `license.lic` file from the MuleSoft Professional Services team. 4. Set the `KUBECONFIG` environment variable to the path to the `kubeconfig` file of the cluster. 5. Verify that the cluster satisfies the prerequisites by running the following command against your `input.yaml` file and the `license.lic` file: `./pcectl pre-install-check --input-file/path/to/input.yaml --license-file /path/to/license.lic` 6. For Anypoint Platform PCE workload to be scheduled on designated nodes, apply the following labels to all worker nodes: - `mulesoft-role=node` - `redis-role=node` 7. Apply the following labels to any three worker nodes. These do not need to be the same three nodes: - `stolon-keeper=yes` - `pce-data-nodes=yes` 8. (Optional) Apply taints to controlplane nodes to ensure that no Anypoint Platform PCE workload is scheduled on the control plane. This also affects any other workload your organization may want to run on control plane nodes. Consult with your Kubernetes cluster admin to perform this step: `mulesoft.pce/controlplane-node=true:NoSchedule` ## Install Anypoint Platform PCE After your infrastructure is correctly set up and you performed a pre-installation verification, MuleSoft Professional Services or your Anypoint PCE Certified Partner will assist you with the installation of Anypoint Platform PCE. To install Anypoint Platform PCE, run this command: `./pcectl install --input-file /path/to/input.yaml --license-file /path/to/license.lic` ## Upgrade Anypoint Platform PCE After you have completed a pre-installation/upgrade verification, MuleSoft Professional Services or your Anypoint PCE Certified Partner will assist you with the upgrade of Anypoint Platform PCE. To upgrade Anypoint Platform PCE, run this command: `./pcectl upgrade /path/to/input.yaml --license-file /path/to/license.lic` ## See Also - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/latest/install-checklist). --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Limiting Concurrent Login in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can limit concurrent logins to Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) so that a user can sign in from only one browser or device at a time. This applies to users in your organization who log in to the platform to use other features (for example, to use Runtime Manager, API Manager, or Access Management). When you limit concurrent logins, a new sign-in from another browser or device signs the user out of the previous session. > [!NOTE] The concurrent login limit configuration isn’t persistent. When you upgrade Anypoint > Platform PCE, the upgrade reverts this configuration. After upgrading, repeat the steps in this > section to restore the concurrent login limit. ## Configure Concurrent Login Limit Run these commands in a terminal where `kubectl` is configured for your Anypoint Platform PCE cluster and you have permissions to operate on it. No specific directory is required. 1. Patch the authentication-server configmap: kubectl patch configmap authentication-server-env -p '{"data":{"AUTH\_SERVER\_FEATURES\_LIMITCONCURRENTSESSION":"true"}}' -n access-management 2. Restart the authentication-server: kubectl rollout restart deployment authentication-server -n access-management The concurrent login limit takes effect after you restart the authentication-server. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Limiting Concurrent Login in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can limit concurrent logins to Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) so that a user can sign in from only one browser or device at a time. This applies to users in your organization who log in to the platform to use other features (for example, to use Runtime Manager, API Manager, or Access Management). When you limit concurrent logins, a new sign-in from another browser or device signs the user out of the previous session. > [!NOTE] The concurrent login limit configuration isn’t persistent. When you upgrade Anypoint > Platform PCE, the upgrade reverts this configuration. After upgrading, repeat the steps in this > section to restore the concurrent login limit. ## Configure Concurrent Login Limit Run these commands in a terminal where `kubectl` is configured for your Anypoint Platform PCE cluster and you have permissions to operate on it. No specific directory is required. 1. Patch the authentication-server configmap: kubectl patch configmap authentication-server-env -p '{"data":{"AUTH\_SERVER\_FEATURES\_LIMITCONCURRENTSESSION":"true"}}' -n access-management 2. Restart the authentication-server: kubectl rollout restart deployment authentication-server -n access-management The concurrent login limit takes effect after you restart the authentication-server. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Limiting Concurrent Login in Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt You can limit concurrent logins to Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) so that a user can sign in from only one browser or device at a time. This applies to users in your organization who log in to the platform to use other features (for example, to use Runtime Manager, API Manager, or Access Management). When you limit concurrent logins, a new sign-in from another browser or device signs the user out of the previous session. > [!NOTE] The concurrent login limit configuration isn’t persistent. When you upgrade Anypoint > Platform PCE, the upgrade reverts this configuration. After upgrading, repeat the steps in this > section to restore the concurrent login limit. ## Configure Concurrent Login Limit Run these commands in a terminal where `kubectl` is configured for your Anypoint Platform PCE cluster and you have permissions to operate on it. No specific directory is required. 1. Patch the authentication-server configmap: kubectl patch configmap authentication-server-env -p '{"data":{"AUTH\_SERVER\_FEATURES\_LIMITCONCURRENTSESSION":"true"}}' -n access-management 2. Restart the authentication-server: kubectl rollout restart deployment authentication-server -n access-management The concurrent login limit takes effect after you restart the authentication-server. ## See Also - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Managing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After your installation is complete, your instance of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) is ready to use. Users registered in your external identity service should have access to your Anypoint platform [organization](https://docs.mulesoft.com/access-management/managing-business-groups). As an admin, you can add [roles](https://docs.mulesoft.com/access-management/roles) to these users. You can also [register servers](https://docs.mulesoft.com/runtime-manager/managing-servers) so that you can then [deploy to them](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers), and so on. Mule app developers in your organization who create Mule apps through the [Anypoint Studio](https://docs.mulesoft.com/studio/) IDE must set up their editors so the Mule apps interface with your locally installed Anypoint Platform rather than the default cloud Anypoint Platform. The following topics describe how to manage Anypoint Platform PCE after installation and configuration is complete: - Close port 61009 in the cluster security group. If you are using AWS, you can use the AWS Management Console to do this. - [Configure Backup and Restore for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/backup-and-disaster-recovery) - [Ops Center](https://docs.mulesoft.com/private-cloud/3.2/managing-via-the-ops-center) - [Add a Product License Using Ops Center](https://docs.mulesoft.com/private-cloud/3.2/ops-center-update-lic) - [Setup Alerts for CPU and Memory Usage](https://docs.mulesoft.com/private-cloud/3.2/config-alerts) - [Allow Addresses for API Platform Proxies](https://docs.mulesoft.com/private-cloud/3.2/config-add-proxy-allowlist) - [Upload Custom Policies and Publish Assets to Exchange Using the Maven Client](https://docs.mulesoft.com/private-cloud/3.2/custom-policies) - [Configure Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/config-studio) - [Update Nodes](https://docs.mulesoft.com/private-cloud/3.2/restarting-a-node) - [Analyze Business and API Data Using ELK](https://docs.mulesoft.com/private-cloud/3.2/ext-analytics-elk) - [Troubleshoot Mule Runtime Engine Registration](https://docs.mulesoft.com/private-cloud/3.2/register-server) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Managing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After your installation is complete, your instance of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) is ready to use. ## Users Using Anypoint Platform PCE Ensure that users registered in your external identity service have access to your Anypoint Platform [organization](https://docs.mulesoft.com/access-management/managing-business-groups). You can also [register servers](https://docs.mulesoft.com/runtime-manager/managing-servers) so that you can then [deploy to them](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers), and so on. ## Mule Developers Using Anypoint Studio Mule app developers in your organization who create Mule apps through the [Anypoint Studio](https://docs.mulesoft.com/studio/) IDE must set up their editors so the Mule apps interface with your locally installed Anypoint Platform rather than the default cloud Anypoint Platform. ## Manage Anypoint Platform PCE The following topics describe how to manage Anypoint Platform PCE: - [Backing up and restoring for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/backup-and-disaster-recovery) - [Allowing addresses for API Platform proxies](https://docs.mulesoft.com/private-cloud/4.0/config-add-proxy-allowlist) - [Uploading custom policies and publish assets to exchange using the Maven client](https://docs.mulesoft.com/private-cloud/4.0/custom-policies) - [Adding Compatible Java Versions for Assets Using the Maven Client](https://docs.mulesoft.com/private-cloud/4.0/adding-compatible-java-versions-using-the-maven-client) - [Configuring Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-studio) - [Analyzing Business and API Data Using ELK](https://docs.mulesoft.com/private-cloud/4.0/operating-about) - [Troubleshooting Mule Runtime Engine Registration](https://docs.mulesoft.com/private-cloud/4.0/register-server) - [Updating Nodes](https://docs.mulesoft.com/private-cloud/4.0/restarting-a-node) ## See Also - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Managing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After your installation is complete, your instance of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) is ready to use. ## Users Using Anypoint Platform PCE Ensure that users registered in your external identity service have access to your Anypoint Platform [organization](https://docs.mulesoft.com/access-management/managing-business-groups). You can also [register servers](https://docs.mulesoft.com/runtime-manager/managing-servers) so that you can then [deploy to them](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers), and so on. ## Mule Developers Using Anypoint Studio Mule app developers in your organization who create Mule apps through the [Anypoint Studio](https://docs.mulesoft.com/studio/) IDE must set up their editors so the Mule apps interface with your locally installed Anypoint Platform rather than the default cloud Anypoint Platform. ## Manage Anypoint Platform PCE The following topics describe how to manage Anypoint Platform PCE: - [Backing up and restoring for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/backup-and-disaster-recovery) - [Allowing addresses for API Platform proxies](https://docs.mulesoft.com/private-cloud/4.1/config-add-proxy-allowlist) - [Uploading custom policies and publish assets to exchange using the Maven client](https://docs.mulesoft.com/private-cloud/4.1/custom-policies) - [Adding Compatible Java Versions for Assets Using the Maven Client](https://docs.mulesoft.com/private-cloud/4.1/adding-compatible-java-versions-using-the-maven-client) - [Configuring Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-studio) - [Analyzing Business and API Data Using ELK](https://docs.mulesoft.com/private-cloud/4.1/operating-about) - [Troubleshooting Mule Runtime Engine Registration](https://docs.mulesoft.com/private-cloud/4.1/register-server) - [Updating Nodes](https://docs.mulesoft.com/private-cloud/4.1/restarting-a-node) ## See Also - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Managing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After your installation is complete, your instance of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) is ready to use. ## Users Using Anypoint Platform PCE Ensure that users registered in your external identity service have access to your Anypoint Platform [organization](https://docs.mulesoft.com/access-management/managing-business-groups). You can also [register servers](https://docs.mulesoft.com/runtime-manager/managing-servers) so that you can then [deploy to them](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers), and so on. ## Mule Developers Using Anypoint Studio Mule app developers in your organization who create Mule apps through the [Anypoint Studio](https://docs.mulesoft.com/studio/) IDE must set up their editors so the Mule apps interface with your locally installed Anypoint Platform rather than the default cloud Anypoint Platform. ## Manage Anypoint Platform PCE The following topics describe how to manage Anypoint Platform PCE: - [Backing up and restoring for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/backup-and-disaster-recovery) - [Allowing addresses for API Platform proxies](https://docs.mulesoft.com/private-cloud/4.2/config-add-proxy-allowlist) - [Uploading custom policies and publish assets to exchange using the Maven client](https://docs.mulesoft.com/private-cloud/4.2/custom-policies) - [Adding Compatible Java Versions for Assets Using the Maven Client](https://docs.mulesoft.com/private-cloud/4.2/adding-compatible-java-versions-using-the-maven-client) - [Configuring Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-studio) - [Analyzing Business and API Data Using ELK](https://docs.mulesoft.com/private-cloud/4.2/operating-about) - [Troubleshooting Mule Runtime Engine Registration](https://docs.mulesoft.com/private-cloud/4.2/register-server) - [Updating Nodes](https://docs.mulesoft.com/private-cloud/4.2/restarting-a-node) ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Managing Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt After your installation is complete, your instance of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) is ready to use. ## Users Using Anypoint Platform PCE Ensure that users registered in your external identity service have access to your Anypoint Platform [organization](https://docs.mulesoft.com/access-management/managing-business-groups). You can also [register servers](https://docs.mulesoft.com/runtime-manager/managing-servers) so that you can then [deploy to them](https://docs.mulesoft.com/runtime-manager/deploying-to-your-own-servers), and so on. ## Mule Developers Using Anypoint Studio Mule app developers in your organization who create Mule apps through the [Anypoint Studio](https://docs.mulesoft.com/studio/) IDE must set up their editors so the Mule apps interface with your locally installed Anypoint Platform rather than the default cloud Anypoint Platform. See [Deploying Mule Applications to Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/deploy-apps). ## Mule Developers Using Anypoint Code Builder Mule app developers in your organization who create Mule apps through [Anypoint Code Builder](https://docs.mulesoft.com/anypoint-code-builder/start-acb-desktop) must connect VS Code to your Anypoint Platform PCE control plane, then package and deploy applications to Runtime Manager. See [Deploying Mule Applications to Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/deploy-apps). ## Manage Anypoint Platform PCE The following topics describe how to manage Anypoint Platform PCE: - [Backing up and restoring for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/backup-and-disaster-recovery) - [Allowing addresses for API Platform proxies](https://docs.mulesoft.com/private-cloud/latest/config-add-proxy-allowlist) - [Uploading custom policies and publish assets to exchange using the Maven client](https://docs.mulesoft.com/private-cloud/latest/custom-policies) - [Adding Compatible Java Versions for Assets Using the Maven Client](https://docs.mulesoft.com/private-cloud/latest/adding-compatible-java-versions-using-the-maven-client) - [Configuring Anypoint Studio for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-studio) - [Deploying Mule Applications to Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/deploy-apps) - [Configuring Anypoint Code Builder for Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-code-builder) - [Deploying Mule Applications Using Anypoint Code Builder](https://docs.mulesoft.com/private-cloud/latest/deploy-apps-code-builder) - [Analyzing Business and API Data Using ELK](https://docs.mulesoft.com/private-cloud/latest/operating-about) - [Troubleshooting Mule Runtime Engine Registration](https://docs.mulesoft.com/private-cloud/latest/register-server) - [Updating Nodes](https://docs.mulesoft.com/private-cloud/latest/restarting-a-node) ## See Also - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/latest/install-workflow) - [Configuring Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/config-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Migrating Anypoint Platform PCE v3.X to v4.X > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] This procedure is also applicable if you need to migrate from PCE 4.x to PCE 4.x instead > of an in-place upgrade. Replace the 3.x system with your source PCE 4.x system in the procedure > below. > [!NOTE] Contact MuleSoft Professional Services or your Anypoint PCE Certified Partner to perform a > migration of your Anypoint Platform PCE. To migrate an Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) deployment from version 3.x to 4.x, perform a full state backup on your existing source cluster (Cluster 1) and restore that data onto a newly provisioned target cluster (Cluster 2) running Anypoint Platform PCE 4.x. Before proceeding, make sure that in your source Anypoint PCE 3.x cluster, you have `ssh` access to your cluster nodes and the necessary `kubectl` permissions for both environments. ## Configure Environment and Security Before generating the backup files, you must correctly identify the platform via DNS and secure it with valid certificates. ### Define DNS and Platform Identification 1. In Amazon Route 53, create a new A or CNAME record to point your load balancer to the PCE 3.x DNS (for example, `pce3.mulesoft.com`). 2. In the Anypoint Platform UI, navigate to **Access Management > DNS/IP** and enter the Platform DNS address hosting your 3.x instance. 3. In the Anypoint Platform UI, click **Save** to apply the DNS changes to the platform configuration. ### Set Up SSL Certificates 1. In your local terminal, generate a 2048-bit RSA key and a self-signed certificate by running the provided `openssl` command. openssl req -x509 -sha256 -nodes -days 365 -newkey rsa:2048 -keyout key.pem -out cert.pem 2. In the Anypoint Platform UI, navigate to **Access Management > Security**. 3. In the Anypoint Platform UI, click **Choose File** for both the **Certificate** and **Private Key** fields to upload the `cert.pem` and `key.pem` files generated in the previous step. ## Back Up Platform State and Infrastructure This section focuses on capturing the database state and service configurations using the Gravity and Kubernetes CLI tools. ### Patch Kubernetes Configuration 1. In the PCE Cluster Terminal, apply the configuration patch to exclude incompatible secrets and buckets from the 4.x migration. kubectl apply -f pce3-update-config.yaml apiVersion: v1 kind: ConfigMap metadata: name: backup-restore-config data: namespaces: "default access-management amf amc arm audit-log api-designer api-manager core-paas design-center exchange mocking mozart pce pce-core visualizer dias monitoring-center" ignore-buckets: "cpc-resources static-assets healthz" ignore-databases: crowdmigrator ^db\\S\*$ sw\_database ignore-secrets: "objectstore-s3-secret telegraf-influxdb-creds nginx-ssl cluster-ca cluster-default-ssl dias-insight-cassandra-connection-password hybrid-rest-cloudhub-keystore hybrid-rest-root-cert runtime-manager-keystore runtime-manager-truststore cloudhub-mcm-keystore" #these hybrid/runtime-manager secrets are ignored since they are generated based on the platform DNS, which is not modified by b/r. See https://github.com/mulesoft/onprem-config-api/blob/master/assets/cloudhub-mcm-keystore/create-keystore.sh ignore-releases: "core-paas-namespaces stolon stolon-amv pce-seaweedfs-app pce-cluster-ssl-app dias-prov-k8s-am-influxdb-comp dias-prov-k8s-am-ingestor-comp" pvc-to-delete: "dias-prov-k8s-cassandra-comp dias-prov-k8s-insight-comp" skip-version-check: "true" apiVersion: v1 data: backup\_s3\_continue\_if\_error: "false" backup\_s3\_max\_attempts: "3" backup\_s3\_sequentially\_bucket\_names: "" backup\_s3\_sleep\_time\_seconds: "30" ignore-buckets: cpc-resources static-assets healthz ignore-databases: crowdmigrator ^db\\S\*$ sw\_database ignore-releases: core-paas-namespaces stolon stolon-amv pce-seaweedfs-app pce-cluster-ssl-app dias-prov-k8s-am-influxdb-comp dias-prov-k8s-am-ingestor-comp ignore-secrets: objectstore-s3-secret telegraf-influxdb-creds nginx-ssl cluster-ca cluster-default-ssl dias-insight-cassandra-connection-password hybrid-rest-cloudhub-keystore hybrid-rest-root-cert runtime-manager-keystore runtime-manager-truststore cloudhub-mcm-keystore namespaces: default access-management amf amc arm audit-log api-designer api-manager core-paas design-center exchange mocking mozart pce pce-core visualizer dias monitoring-center pithos\_pod\_count: "2" pvc-to-delete: dias-prov-k8s-cassandra-comp dias-prov-k8s-insight-comp restore\_in\_parallel: "false" restore\_s3\_max\_attempts: "3" restore\_s3\_sleep\_time\_seconds: "30" skip-version-check: "true" kind: ConfigMap metadata: name: backup-restore-config ### Execute Gravity System Backup 1. In the PCE Cluster Terminal (Backup Node), trigger the core platform state backup by executing the gravity backup command. gravity backup /var/lib/gravity/planet/share/backup-3x-$(date +%d-%m-%y).tar.gz 2. In the PCE Cluster Terminal (Backup Node), generate the Anypoint Monitoring and Visualizer (AMV) specific backup using the internal restore script. /var/lib/gravity/site/packages/unpacked/gravitational.io/anypoint/3.2.0\*/resources/kubernetes/amv-backup-restore/amv-backup-restore.sh ## Preserve Persistent Data (NFS/EFS) In addition to the database state, you must manually preserve the project files and working directories stored on the distributed file system. ### Mount the External Storage 1. In the PCE Cluster Terminal, create a mount point and install the necessary NFS utilities. mkdir nfs\_mount\_folder sudo yum install nfs-utils -y 2. In the PCE Cluster Terminal, mount your AWS EFS (or external NFS) to the local directory. sudo mount -t nfs4 -o rw,vers=4.1 :/ /home/ec2-user/nfs\_mount\_folder ### Compress and Offload Assets 1. In the PCE Cluster Terminal, navigate to the mounted folder and compress the project and working directories. tar -cvzf projects.tar.gz projects tar -cvzf working\_Directories.tar.gz working\_Directories 2. In your Local Terminal, securely download the four critical artifacts (Platform Backup, AMV Backup, Projects Archive, and Working Directories Archive) to your local machine. scp -i your-key.pem ec2-user@:/tmp/backup-files/\*.tar.gz ~/Downloads/pce\_migration\_assets/ ## Initialize the Target 4.x Environment Before restoring data, you must configure the new cluster to recognize the platform identity and security parameters. ### Update Platform Routing and Identification 1. In Amazon Route 53, update your DNS records to point the load balancer to the new Anypoint PCE 4.x cluster (Cluster 2). 2. In the Anypoint Platform UI (New Cluster), navigate to **Access Management > DNS/IP** and enter the new DNS address assigned to your PCE 4.x instance. ### Apply New Security Certificates 1. In the Anypoint Platform UI (New Cluster), navigate to **Access Management > Security** and upload the certificates generated for the 4.x environment. 2. In the PCE 4.x Cluster Terminal, monitor the cluster to ensure all jobs related to DNS and certificate updates complete successfully and that all platform pods are in a `Running` state. ## Transfer and Stage Migration Assets Move the data extracted from PCE 3.x into the target 4.x infrastructure. ### Mount the Target External Storage 1. In the PCE 4.x Cluster Terminal, create the `nfs_mount_folder` and ensure the `nfs-utils` package is installed on the node. 2. In the PCE 4.x Cluster Terminal, mount your AWS EFS or external storage to the local mount point using the `mount -t nfs4` command. ### Upload and Extract Backup Data 1. In your Local Terminal, transfer the backup archives and project directories from your local machine to the new EC2 instance using the `scp` command. 2. In the PCE 4.x Cluster Terminal, extract the `projects.tar.gz` and `working_Directories.tar.gz` files into their respective persistent storage paths. ## Execute the Restoration Process Finalize the migration by updating system flags and triggering the restoration services via API. ### Adjust Kubernetes Migration Settings 1. In the PCE 4.x Cluster Terminal, open the backup-restore configuration for editing by running the `kubectl edit` command. kubectl edit cm backup-restore-config -n pce-core 2. In the editor, update the `skip-version-check` property to true to authorize the version-jump migration. ### Invoke Restoration Services via API 1. In your Local Terminal, obtain an authentication bearer token by providing your credentials to the login endpoint. curl -k --request POST \\ --header "Content-Type: application/json" \\ --data '{"username":"", "password": ""}' \\ https:///accounts/login 2. In your Local Terminal, trigger the core platform restoration using the backup file and target NFS server details. curl -k https:///platform/restore -X POST \\ -H "Authorization: Bearer ${token}" \\ -H "Content-Type: application/json" \\ -d "{\\"backup-file-name\\": \\"${backupFileName}.tar.gz\\", \\"nfs-server\\": \\"${fileSystemDns}\\", \\"nfs-path\\": \\"/\\"}" 3. In your Local Terminal, trigger the AMV restoration to migrate Monitoring and Visualizer data. curl -k https:///platform/amv/restore -X POST \\ -H "Authorization: Bearer ${token}" \\ -H "Content-Type: application/json" \\ -d "{\\"backup-file-name\\": \\"${backupFileName}.tar.gz\\", \\"nfs-server\\": \\"${fileSystemDns}\\", \\"nfs-path\\": \\"/\\"}" ## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.2/) - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Migrating Anypoint Platform PCE v3.X to v4.X > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] This procedure is also applicable if you need to migrate from PCE 4.x to PCE 4.x instead > of an in-place upgrade. Replace the 3.x system with your source PCE 4.x system in the procedure > below. > [!NOTE] Contact MuleSoft Professional Services or your Anypoint PCE Certified Partner to perform a > migration of your Anypoint Platform PCE. To migrate an Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) deployment from version 3.x to 4.x, perform a full state backup on your existing source cluster (Cluster 1) and restore that data onto a newly provisioned target cluster (Cluster 2) running Anypoint Platform PCE 4.x. Before proceeding, make sure that in your source Anypoint PCE 3.x cluster, you have `ssh` access to your cluster nodes and the necessary `kubectl` permissions for both environments. ## Configure Environment and Security Before generating the backup files, you must correctly identify the platform via DNS and secure it with valid certificates. ### Define DNS and Platform Identification 1. In Amazon Route 53, create a new A or CNAME record to point your load balancer to the PCE 3.x DNS (for example, `pce3.mulesoft.com`). 2. In the Anypoint Platform UI, navigate to **Access Management > DNS/IP** and enter the Platform DNS address hosting your 3.x instance. 3. In the Anypoint Platform UI, click **Save** to apply the DNS changes to the platform configuration. ### Set Up SSL Certificates 1. In your local terminal, generate a 2048-bit RSA key and a self-signed certificate by running the provided `openssl` command. openssl req -x509 -sha256 -nodes -days 365 -newkey rsa:2048 -keyout key.pem -out cert.pem 2. In the Anypoint Platform UI, navigate to **Access Management > Security**. 3. In the Anypoint Platform UI, click **Choose File** for both the **Certificate** and **Private Key** fields to upload the `cert.pem` and `key.pem` files generated in the previous step. ## Back Up Platform State and Infrastructure This section focuses on capturing the database state and service configurations using the Gravity and Kubernetes CLI tools. ### Patch Kubernetes Configuration 1. In the PCE Cluster Terminal, apply the configuration patch to exclude incompatible secrets and buckets from the 4.x migration. kubectl apply -f pce3-update-config.yaml apiVersion: v1 kind: ConfigMap metadata: name: backup-restore-config data: namespaces: "default access-management amf amc arm audit-log api-designer api-manager core-paas design-center exchange mocking mozart pce pce-core visualizer dias monitoring-center" ignore-buckets: "cpc-resources static-assets healthz" ignore-databases: crowdmigrator ^db\\S\*$ sw\_database ignore-secrets: "objectstore-s3-secret telegraf-influxdb-creds nginx-ssl cluster-ca cluster-default-ssl dias-insight-cassandra-connection-password hybrid-rest-cloudhub-keystore hybrid-rest-root-cert runtime-manager-keystore runtime-manager-truststore cloudhub-mcm-keystore" #these hybrid/runtime-manager secrets are ignored since they are generated based on the platform DNS, which is not modified by b/r. See https://github.com/mulesoft/onprem-config-api/blob/master/assets/cloudhub-mcm-keystore/create-keystore.sh ignore-releases: "core-paas-namespaces stolon stolon-amv pce-seaweedfs-app pce-cluster-ssl-app dias-prov-k8s-am-influxdb-comp dias-prov-k8s-am-ingestor-comp" pvc-to-delete: "dias-prov-k8s-cassandra-comp dias-prov-k8s-insight-comp" skip-version-check: "true" apiVersion: v1 data: backup\_s3\_continue\_if\_error: "false" backup\_s3\_max\_attempts: "3" backup\_s3\_sequentially\_bucket\_names: "" backup\_s3\_sleep\_time\_seconds: "30" ignore-buckets: cpc-resources static-assets healthz ignore-databases: crowdmigrator ^db\\S\*$ sw\_database ignore-releases: core-paas-namespaces stolon stolon-amv pce-seaweedfs-app pce-cluster-ssl-app dias-prov-k8s-am-influxdb-comp dias-prov-k8s-am-ingestor-comp ignore-secrets: objectstore-s3-secret telegraf-influxdb-creds nginx-ssl cluster-ca cluster-default-ssl dias-insight-cassandra-connection-password hybrid-rest-cloudhub-keystore hybrid-rest-root-cert runtime-manager-keystore runtime-manager-truststore cloudhub-mcm-keystore namespaces: default access-management amf amc arm audit-log api-designer api-manager core-paas design-center exchange mocking mozart pce pce-core visualizer dias monitoring-center pithos\_pod\_count: "2" pvc-to-delete: dias-prov-k8s-cassandra-comp dias-prov-k8s-insight-comp restore\_in\_parallel: "false" restore\_s3\_max\_attempts: "3" restore\_s3\_sleep\_time\_seconds: "30" skip-version-check: "true" kind: ConfigMap metadata: name: backup-restore-config ### Execute Gravity System Backup 1. In the PCE Cluster Terminal (Backup Node), trigger the core platform state backup by executing the gravity backup command. gravity backup /var/lib/gravity/planet/share/backup-3x-$(date +%d-%m-%y).tar.gz 2. In the PCE Cluster Terminal (Backup Node), generate the Anypoint Monitoring and Visualizer (AMV) specific backup using the internal restore script. /var/lib/gravity/site/packages/unpacked/gravitational.io/anypoint/3.2.0\*/resources/kubernetes/amv-backup-restore/amv-backup-restore.sh ## Preserve Persistent Data (Shared Filesystem) In addition to the database state, you must manually preserve the project files and working directories stored on the platform shared filesystem. ### Mount the External Storage 1. In the PCE Cluster Terminal, create a mount point and install the necessary NFS utilities. mkdir nfs\_mount\_folder sudo yum install nfs-utils -y 2. In the PCE Cluster Terminal, mount your AWS EFS (or external NFS) to the local directory. sudo mount -t nfs4 -o rw,vers=4.1 :/ /home/ec2-user/nfs\_mount\_folder ### Compress and Offload Assets 1. In the PCE Cluster Terminal, navigate to the mounted folder and compress the project and working directories. tar -cvzf projects.tar.gz projects tar -cvzf working\_Directories.tar.gz working\_Directories 2. In your Local Terminal, securely download the four critical artifacts (Platform Backup, AMV Backup, Projects Archive, and Working Directories Archive) to your local machine. scp -i your-key.pem ec2-user@:/tmp/backup-files/\*.tar.gz ~/Downloads/pce\_migration\_assets/ ## Initialize the Target 4.x Environment Before restoring data, you must configure the new cluster to recognize the platform identity and security parameters. ### Update Platform Routing and Identification 1. In Amazon Route 53, update your DNS records to point the load balancer to the new Anypoint PCE 4.x cluster (Cluster 2). 2. In the Anypoint Platform UI (New Cluster), navigate to **Access Management > DNS/IP** and enter the new DNS address assigned to your PCE 4.x instance. ### Apply New Security Certificates 1. In the Anypoint Platform UI (New Cluster), navigate to **Access Management > Security** and upload the certificates generated for the 4.x environment. 2. In the PCE 4.x Cluster Terminal, monitor the cluster to ensure all jobs related to DNS and certificate updates complete successfully and that all platform pods are in a `Running` state. ## Transfer and Stage Migration Assets Move the data extracted from PCE 3.x into the target 4.x infrastructure. ### Mount the Target External Storage 1. In the PCE 4.x Cluster Terminal, create the `nfs_mount_folder` and ensure the `nfs-utils` package is installed on the node. 2. In the PCE 4.x Cluster Terminal, mount your AWS EFS or external storage to the local mount point using the `mount -t nfs4` command. ### Upload and Extract Backup Data 1. In your Local Terminal, transfer the backup archives and project directories from your local machine to the new EC2 instance using the `scp` command. 2. In the PCE 4.x Cluster Terminal, extract the `projects.tar.gz` and `working_Directories.tar.gz` files into their respective persistent storage paths. ## Execute the Restoration Process Finalize the migration by updating system flags and triggering the restoration services via API. ### Adjust Kubernetes Migration Settings 1. In the PCE 4.x Cluster Terminal, open the backup-restore configuration for editing by running the `kubectl edit` command. kubectl edit cm backup-restore-config -n pce-core 2. In the editor, update the `skip-version-check` property to true to authorize the version-jump migration. ### Invoke Restoration Services via API 1. In your Local Terminal, obtain an authentication bearer token by providing your credentials to the login endpoint. curl -k --request POST \\ --header "Content-Type: application/json" \\ --data '{"username":"", "password": ""}' \\ https:///accounts/login 2. In your Local Terminal, trigger the core platform restoration using the backup file and target NFS server details. curl -k https:///platform/restore -X POST \\ -H "Authorization: Bearer ${token}" \\ -H "Content-Type: application/json" \\ -d "{\\"backup-file-name\\": \\"${backupFileName}.tar.gz\\", \\"nfs-server\\": \\"${fileSystemDns}\\", \\"nfs-path\\": \\"/\\"}" 3. In your Local Terminal, trigger the AMV restoration to migrate Monitoring and Visualizer data. curl -k https:///platform/amv/restore -X POST \\ -H "Authorization: Bearer ${token}" \\ -H "Content-Type: application/json" \\ -d "{\\"backup-file-name\\": \\"${backupFileName}.tar.gz\\", \\"nfs-server\\": \\"${fileSystemDns}\\", \\"nfs-path\\": \\"/\\"}" ## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/latest/) - [Installing or Upgrading Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/latest/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Network Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To ensure the performance and stability of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), every node in your Anypoint Platform PCE environment must meet the network requirements described in this topic. > [!IMPORTANT] Before you install Anypoint Platform PCE, your infrastructure team must review each > of the following sections and verify that your environment meets the stated network requirements. > If needed, contact your MuleSoft representative for assistance. ## Verify Static IPs All nodes in the cluster must have static private IPv4 assigned to them, and these must persist after restarts. If IP addresses are not persistent between reboots, the cluster can fail. ## Verify VXLAN The version of Kubernetes supported by Anypoint Platform PCE uses an overlay VXLAN and UDP transport to encapsulate traffic. There is direct communication between cluster components. The following table lists the ports used for inter-host communication. The table uses the following source and destination descriptions: - Anypoint Platform PCE nodes: Nodes that are members of the cluster. - Installer node: The node on which you started the installation. - Load Balancer: Any source that is outside the cluster. - localhost: The port that is used only within the host where the request started. You must ensure that all of the following ports are configured correctly on all nodes.
Table 1. Port Configuration Requirements
ProtocolPort/RangePurposeSourceDestination

TCP/UDP

53

Internal cluster DNS

localhost

localhost

TCP

2379, 2380, 4001, 7001

etcd distributed database

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

4242

Installer

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

6443

Kubernetes API server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9443

HAProxy to load balance Kubernetes API Server

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

9100

Prometheus node-exporter

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10248, 10250

Kubernetes Kubelet

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10249, 10256

Kubernetes kube-proxy

Anypoint Platform PCE

Anypoint Platform PCE nodes

TCP

10251-10252

Kubernetes controller-manager and scheduler

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

10257-10259

Kubernetes controller-manager and scheduler (secure)

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

5000

Docker registry

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

3009-3012

Cluster control plane

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

3022

Cluster control plane

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

3023-3025

Cluster control plane

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

7575

Cluster control plane

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

7580

Cluster control plane monitoring endpoint

localhost

localhost

TCP

6060

Cluster control plane profiling endpoint

localhost

localhost

TCP

30000-32767

Internal services port range

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

61009-61010

Installer port ranges (only used during install)

Anypoint Platform PCE nodes

installer node

TCP

61022-61024

Installer port ranges (only used during install)

Anypoint Platform PCE nodes

installer node

TCP

61025

Installer port ranges (only used during install)

localhost

installer node

UDP

8472

Overlay VXLAN network

Anypoint Platform PCE nodes

Anypoint Platform PCE nodes

TCP

32009

Ops Center Admin UI

Load Balancer

Anypoint Platform PCE nodes

TCP

30080

HTTP public access

Load Balancer

Anypoint Platform PCE nodes

TCP

30443

HTTPS public access

Load Balancer

Anypoint Platform PCE nodes

TCP

30889

Mule Runtime Websocket

Load Balancer

Anypoint Platform PCE nodes

TCP

30883

Mule Runtime Authentication Service (for certificate renewals)

Load Balancer

Anypoint Platform PCE nodes

You can use netcat or a similar utility to verify port configuration. Anypoint Platform PCE nodes need to hit the platform API address on the load balancer via HTTPS. ## Enable Required Ports - You must enable TCP ports 53, 3011, and 3012 on all nodes to allow communication with the database cluster. - If you plan to use the GUI installer to install Anypoint Platform PCE, you must enable port 61009. ## Verify NAT Traffic Requirements In some situations, the Kubernetes overlay network uses NAT. NAT requires that nodes be able to send and receive packages with a source and destination different from the node’s internal IP. Any software that monitors the network traffic must be tuned to allow this type of network interaction. ## Disable OS NetworkManager Services In some operating systems, like RHEL/CentOS 8.4, NetworkManager includes two extra services: `nm-cloud-setup.service` and `nm-cloud-setup.timer`. These services add routing tables that interfere with the configuration of the Kubernetes Container Network Interface (CNI) plugin. If those services exist, disable them and restart the node before proceeding with the installation. ## Verify SMTP Server Requirements Your network must include an SMTP server to manage email alerts that are triggered by the platform. Anypoint Platform PCE must be connected to this SMTP server. For additional information, refer to the instructions in [Configure SMTP in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/access-management-SMTP). ## Verify External Identity Provider Configuration and Client Management Your network must include an external identity provider that is compatible with LDAP, OpenIDConnect, or SAML. To configure an External Identity Provider for user SSO (LDAP, OpenIDConnect, and SAML) or for external client management as OpenID Connect, the cluster must be able to access the external identity provider endpoint. For additional information, refer to the instructions in [Configure an External Identity Provider in Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/install-config-ldap-pce). ## Synchronize Clocks Ensure that the clocks on all servers are synchronized. Unsynchronized clocks can cause problems with platform communications. ## See Also - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-platform) - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # NFS Server Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), > your infrastructure team must review each of the following sections and verify that your > environment meets the stated requirements. If needed, contact your MuleSoft representative for > assistance to ensure your NFS server is configured correctly. ## Verify NFS Server Setup Anypoint Platform PCE version 2.0 and later requires a Network File System (NFS) server and an available NFS endpoint. Your NFS server must meet the following requirements: - Must be running on a filesystem that has at least 250 GB available. - Must use any of these supported NFS versions: 4.0 or 4.1 - Can be backed by a network-attached storage (NAS) system. - Must have port 2049 enabled for both TCP and UDP. - Configure your network so that all of the PCE nodes can access the NFS server. The direct NFS client is Kubernetes 1.9 and requires conformance with NFS version 4. NFS 4.0 and 4.1 don’t use the `portmapper` service. ## Verify Required Port Is Open The NFS server requires port 2049 to be open for TCP and UDP. On each node, use netcat or a similar utility to verify that the required port is open. Netcat is a computer networking utility used to read and write to network connections using TCP or UDP. On each node, run: `nc -zv 2049` > [!NOTE] By default, the netcat tool is not included with Anypoint Platform PCE. If you receive a > `command not found` error, you can download and install netcat or a similar utility. When the responses from netcat are successful, you receive a response similar to the following example: \[ec2-user@ip-10-1-1-97 ~\]$ nc -zv 2049 Ncat: Version 7.50 ( https://nmap.org/ncat ) Ncat: Connected to 10.1.0.155:2049. Ncat: 0 bytes sent, 0 bytes received in 0.01 seconds. ## Verify NFS Server Is Mounted On each node, mount the NFS server in a temporary directory. 1. Switch to root: `sudo su` 2. To mount the NFS in a temporary directory, run the following commands and replace both the `` and the `` with your own values: `mkdir -p /mnt/home` `mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/` If the mount fails, you do not have the required permissions. ## Verify NFS Performance Perform careful analysis of your environment, both from the client and server point of view for optimal NFS performance. This involves: - Testing the performance and latency of your NFS server. Specifically, these operations will transfer data from the client (Anypoint Platform PCE environment) to the NFS server and measure how long each transfer takes. - Using the command `time dd if=/dev/zero of=/mnt/home/ bs= count=` to test NFS server performance while transferring files and specifying the following parameters: - Names `` - Block size and number of blocks in the transfer: `` and ``, respectively. - Validating how Anypoint Platform PCE handles NFS file transfers. Perform read and write operations with both large (128MB) and small (80KB) files. Perform all tests on each node. ### Test Writing and Reading Large Files on the NFS Server Perform the following read and write tests using large files: - File size: 128 MB - Block size: 4 KB - Number of blocks: 32768 While the tests are running, different files are generated (one file per test). 1. Create a directory in which to mount the NFS server: sudo su mkdir -p /mnt/home 2. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform writes (five files of 128 MB each) on the NFS server: time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done The command output is similar to the following: \[root@ip-10-1-1-97 ec2-user\]# time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done 32768+0 records in 32768+0 records out 134217728 bytes (134 MB) copied, 2.12805 s, 63.1 MB/s . . . real 0m8.378s user 0m0.034s sys 0m0.792s The `real` parameter in the test output should be less than 15 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount /mnt/home 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server by running the following command: time for i in {1..5}; do dd if=/mnt/home/greatfile$i.test of=/dev/null bs=4k; done The `real` parameter in the test output should be less than 15 seconds. #### Clear the Caches Unmount the NFS server to clear out any caches: umount /mnt/home ### Test Writing and Reading Small Files on the NFS Server Test small files as you did large files but with these sizes: - File size: 80 KB - Block size: 4 KB - Number of blocks: 20 Before starting the test, verify that the `/mnt/home` directory from the previous test is created. 1. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 2. Perform writes (five files of 80 KB each) on the NFS server: time for i in {1..30}; do dd if=/dev/zero of=/mnt/home/smallfile$i.test bs=4k count=20; done The `real` parameter in the test output should be around 2 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server: time for i in {1..30}; do dd if=/mnt/home/smallfile$i.test of=/dev/null bs=4k; done The important parameter in the test output is the `real` parameter, which should be around 2 seconds. #### Delete the Test Files When you finish the performance testing, delete the test files. 1. Remove all generated files by running: rm -f /mnt/home/\*.test 2. Unmount the NFS server: umount /mnt/home ## See Also - [Troubleshoot NFS](https://docs.mulesoft.com/private-cloud/3.2/troubleshoot-nfs) - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-platform) - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # NFS Server Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), > your infrastructure team must review each of the following sections and verify that your > environment meets the stated requirements. If needed, contact your MuleSoft representative for > assistance to ensure your NFS server is configured correctly. ## Verify NFS Server Setup Anypoint Platform PCE version 2.0 and later requires a Network File System (NFS) server and an available NFS endpoint. Your NFS server must meet the following requirements: - Must be running on a filesystem that has at least 250 GB available. - Must use any of these supported NFS versions: 4.0 or 4.1 - Can be backed by a network-attached storage (NAS) system. - Must have port 2049 enabled for both TCP and UDP. - Configure your network so that all of the PCE nodes can access the NFS server. The direct NFS client is Kubernetes 1.9 and requires conformance with NFS version 4. NFS 4.0 and 4.1 do not use the `portmapper` service. ## Verify Required Port Is Open The NFS server requires port 2049 to be open for TCP and UDP. On each node, use netcat or a similar utility to verify that the required port is open. Netcat is a computer networking utility used to read and write to network connections using TCP or UDP. On each node, run: `nc -zv 2049` > [!NOTE] By default, the netcat tool is not included with Anypoint Platform PCE. If you receive a > `command not found` error, you can download and install netcat or a similar utility. When the responses from netcat are successful, you receive a response similar to the following example: \[ec2-user@ip-10-1-1-97 ~\]$ nc -zv 2049 Ncat: Version 7.50 ( https://nmap.org/ncat ) Ncat: Connected to 10.1.0.155:2049. Ncat: 0 bytes sent, 0 bytes received in 0.01 seconds. ## Verify NFS Server Is Mounted On each node, mount the NFS server in a temporary directory. 1. Switch to root: `sudo su` 2. To mount the NFS in a temporary directory, run the following commands and replace both the `` and the `` with your own values: `mkdir -p /mnt/home` `mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/` If the mount fails, you do not have the required permissions. ## Verify NFS Performance Perform careful analysis of your environment, both from the client and server point of view for optimal NFS performance. This involves: - Testing the performance and latency of your NFS server. Specifically, these operations will transfer data from the client (Anypoint Platform PCE environment) to the NFS server and measure how long each transfer takes. - Using the command `time dd if=/dev/zero of=/mnt/home/ bs= count=` to test NFS server performance while transferring files and specifying the following parameters: - Names `` - Block size and number of blocks in the transfer: `` and ``, respectively. - Validating how Anypoint Platform PCE handles NFS file transfers. Perform read and write operations with both large (128MB) and small (80KB) files. Perform all tests on each node. ### Test Writing and Reading Large Files on the NFS Server Perform the following read and write tests using large files: - File size: 128 MB - Block size: 4 KB - Number of blocks: 32768 While the tests are running, different files are generated (one file per test). 1. Create a directory in which to mount the NFS server: sudo su mkdir -p /mnt/home 2. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform writes (five files of 128 MB each) on the NFS server: time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done The command output is similar to the following: \[root@ip-10-1-1-97 ec2-user\]# time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done 32768+0 records in 32768+0 records out 134217728 bytes (134 MB) copied, 2.12805 s, 63.1 MB/s . . . real 0m8.378s user 0m0.034s sys 0m0.792s The `real` parameter in the test output should be less than 15 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount /mnt/home 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server by running the following command: time for i in {1..5}; do dd if=/mnt/home/greatfile$i.test of=/dev/null bs=4k; done The `real` parameter in the test output should be less than 15 seconds. #### Clear the Caches Unmount the NFS server to clear out any caches: umount /mnt/home ### Test Writing and Reading Small Files on the NFS Server Test small files as you did large files but with these sizes: - File size: 80 KB - Block size: 4 KB - Number of blocks: 20 Before starting the test, verify that the `/mnt/home` directory from the previous test is created. 1. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 2. Perform writes (five files of 80 KB each) on the NFS server: time for i in {1..30}; do dd if=/dev/zero of=/mnt/home/smallfile$i.test bs=4k count=20; done The `real` parameter in the test output should be around 2 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server: time for i in {1..30}; do dd if=/mnt/home/smallfile$i.test of=/dev/null bs=4k; done The important parameter in the test output is the `real` parameter, which should be around 2 seconds. #### Delete the Test Files When you finish the performance testing, delete the test files. 1. Remove all generated files by running: rm -f /mnt/home/\*.test 2. Unmount the NFS server: umount /mnt/home ## See Also - [Troubleshoot NFS](https://docs.mulesoft.com/private-cloud/4.0/troubleshoot-nfs) - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # NFS Server Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), > your infrastructure team must review each of the following sections and verify that your > environment meets the stated requirements. If needed, contact your MuleSoft representative for > assistance to ensure your NFS server is configured correctly. ## Verify NFS Server Setup Anypoint Platform PCE version 2.0 and later requires a Network File System (NFS) server and an available NFS endpoint. Your NFS server must meet the following requirements: - Must be running on a filesystem that has at least 250 GB available. - Must be using NFS 4.x. - Can be backed by a network-attached storage (NAS) system. - Must have port 2049 enabled for both TCP and UDP. - Configure your network so that all of the PCE nodes can access the NFS server. The direct NFS client is Kubernetes 1.9 and requires conformance with NFS version 4. NFS 4.x does not use the `portmapper` service. ## Verify Required Port Is Open The NFS server requires port 2049 to be open for TCP and UDP. On each node, use netcat or a similar utility to verify that the required port is open. Netcat is a computer networking utility used to read and write to network connections using TCP or UDP. On each node, run: `nc -zv 2049` > [!NOTE] By default, the netcat tool is not included with Anypoint Platform PCE. If you receive a > `command not found` error, you can download and install netcat or a similar utility. When the responses from netcat are successful, you receive a response similar to the following example: \[ec2-user@ip-10-1-1-97 ~\]$ nc -zv 2049 Ncat: Version 7.50 ( https://nmap.org/ncat ) Ncat: Connected to 10.1.0.155:2049. Ncat: 0 bytes sent, 0 bytes received in 0.01 seconds. ## Verify NFS Server Is Mounted On each node, mount the NFS server in a temporary directory. 1. Switch to root: `sudo su` 2. To mount the NFS in a temporary directory, run the following commands and replace both the `` and the `` with your own values: `mkdir -p /mnt/home` `mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/` If the mount fails, you do not have the required permissions. ## Verify NFS Performance Perform careful analysis of your environment, both from the client and server point of view for optimal NFS performance. This involves: - Testing the performance and latency of your NFS server. Specifically, these operations will transfer data from the client (Anypoint Platform PCE environment) to the NFS server and measure how long each transfer takes. - Using the command `time dd if=/dev/zero of=/mnt/home/ bs= count=` to test NFS server performance while transferring files and specifying the following parameters: - Names `` - Block size and number of blocks in the transfer: `` and ``, respectively. - Validating how Anypoint Platform PCE handles NFS file transfers. Perform read and write operations with both large (128MB) and small (80KB) files. Perform all tests on each node. ### Test Writing and Reading Large Files on the NFS Server Perform the following read and write tests using large files: - File size: 128 MB - Block size: 4 KB - Number of blocks: 32768 While the tests are running, different files are generated (one file per test). 1. Create a directory in which to mount the NFS server: sudo su mkdir -p /mnt/home 2. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform writes (five files of 128 MB each) on the NFS server: time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done The command output is similar to the following: \[root@ip-10-1-1-97 ec2-user\]# time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done 32768+0 records in 32768+0 records out 134217728 bytes (134 MB) copied, 2.12805 s, 63.1 MB/s . . . real 0m8.378s user 0m0.034s sys 0m0.792s The `real` parameter in the test output should be less than 15 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount /mnt/home 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server by running the following command: time for i in {1..5}; do dd if=/mnt/home/greatfile$i.test of=/dev/null bs=4k; done The `real` parameter in the test output should be less than 15 seconds. #### Clear the Caches Unmount the NFS server to clear out any caches: umount /mnt/home ### Test Writing and Reading Small Files on the NFS Server Test small files as you did large files but with these sizes: - File size: 80 KB - Block size: 4 KB - Number of blocks: 20 Before starting the test, verify that the `/mnt/home` directory from the previous test is created. 1. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 2. Perform writes (five files of 80 KB each) on the NFS server: time for i in {1..30}; do dd if=/dev/zero of=/mnt/home/smallfile$i.test bs=4k count=20; done The `real` parameter in the test output should be around 2 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server: time for i in {1..30}; do dd if=/mnt/home/smallfile$i.test of=/dev/null bs=4k; done The important parameter in the test output is the `real` parameter, which should be around 2 seconds. #### Delete the Test Files When you finish the performance testing, delete the test files. 1. Remove all generated files by running: rm -f /mnt/home/\*.test 2. Unmount the NFS server: umount /mnt/home ## See Also - [Troubleshoot NFS](https://docs.mulesoft.com/private-cloud/4.1/troubleshoot-nfs) - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # NFS Server Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), > your infrastructure team must review each of the following sections and verify that your > environment meets the stated requirements. If needed, contact your MuleSoft representative for > assistance to ensure your NFS server is configured correctly. ## CSI Driver Requirement Starting with Anypoint Platform PCE 4.2.0, a pre-installed Container Storage Interface (CSI) driver is a mandatory requirement. Anypoint Platform PCE uses the Kubernetes CSI driver to mount NFS volumes into pods, replacing the previous method that relied on host binaries. Some key benefits: - Enhanced support for environments where custom tools cannot be installed on host nodes. - Direct support for Amazon EKS users who can only use the Amazon EFS CSI Driver. ### Configure the CSI Driver In the `platformConfiguration` section of your `input.yaml` file, specify the CSI driver name: ```yaml platformConfiguration: fileSystemCsiDriverName: "" # CSI driver name (e.g., nfs.csi.k8s.io, efs.csi.aws.com) ``` ### EFS CSI Driver Configuration When using the Amazon EFS CSI Driver (for example, for Amazon EKS with EFS installations), provide four individual CSI volume handles. These volume handles must point to four access points that all target the same directory in the EFS: ```yaml platformConfiguration: fileSystemCsiDriverName: "efs.csi.aws.com" fileSystemWcCsiVolumeHandle: "" # CSI volume handle for wc filesystem fileSystem01CsiVolumeHandle: "" # CSI volume handle for 01 filesystem fileSystemBareCsiVolumeHandle: "" # CSI volume handle for bare filesystem fileSystemBackupRestoreCsiVolumeHandle: "" # CSI volume handle for backup-restore filesystem ``` Each volume handle value follows the format `::`, for example: `fs-0835b0d1eb7588eb1::fsap-03f6c68563ceb5acf`. > [!NOTE] When using the EFS CSI Driver, you do not need to provide `fileSystemDNS` or > `fileSystemPath` values. ## Verify NFS Server Setup Anypoint Platform PCE version 2.0 and later requires a Network File System (NFS) server and an available NFS endpoint. Your NFS server must meet the following requirements: - Must be running on a filesystem that has at least 250 GB available. - Must be using NFS 4.x. - Can be backed by a network-attached storage (NAS) system. - Must have port 2049 enabled for both TCP and UDP. - Configure your network so that all of the PCE nodes can access the NFS server. The direct NFS client is Kubernetes 1.9 and requires conformance with NFS version 4. NFS 4.x does not use the `portmapper` service. ## Verify Required Port Is Open The NFS server requires port 2049 to be open for TCP and UDP. On each node, use netcat or a similar utility to verify that the required port is open. Netcat is a computer networking utility used to read and write to network connections using TCP or UDP. On each node, run: `nc -zv 2049` > [!NOTE] By default, the netcat tool is not included with Anypoint Platform PCE. If you receive a > `command not found` error, you can download and install netcat or a similar utility. When the responses from netcat are successful, you receive a response similar to the following example: \[ec2-user@ip-10-1-1-97 ~\]$ nc -zv 2049 Ncat: Version 7.50 ( https://nmap.org/ncat ) Ncat: Connected to 10.1.0.155:2049. Ncat: 0 bytes sent, 0 bytes received in 0.01 seconds. ## Verify NFS Server Is Mounted On each node, mount the NFS server in a temporary directory. 1. Switch to root: `sudo su` 2. To mount the NFS in a temporary directory, run the following commands and replace both the `` and the `` with your own values: `mkdir -p /mnt/home` `mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/` If the mount fails, you do not have the required permissions. ## Verify NFS Performance Perform careful analysis of your environment, both from the client and server point of view for optimal NFS performance. This involves: - Testing the performance and latency of your NFS server. Specifically, these operations will transfer data from the client (Anypoint Platform PCE environment) to the NFS server and measure how long each transfer takes. - Using the command `time dd if=/dev/zero of=/mnt/home/ bs= count=` to test NFS server performance while transferring files and specifying the following parameters: - Names `` - Block size and number of blocks in the transfer: `` and ``, respectively. - Validating how Anypoint Platform PCE handles NFS file transfers. Perform read and write operations with both large (128MB) and small (80KB) files. Perform all tests on each node. ### Test Writing and Reading Large Files on the NFS Server Perform the following read and write tests using large files: - File size: 128 MB - Block size: 4 KB - Number of blocks: 32768 While the tests are running, different files are generated (one file per test). 1. Create a directory in which to mount the NFS server: sudo su mkdir -p /mnt/home 2. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform writes (five files of 128 MB each) on the NFS server: time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done The command output is similar to the following: \[root@ip-10-1-1-97 ec2-user\]# time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done 32768+0 records in 32768+0 records out 134217728 bytes (134 MB) copied, 2.12805 s, 63.1 MB/s . . . real 0m8.378s user 0m0.034s sys 0m0.792s The `real` parameter in the test output should be less than 15 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount /mnt/home 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server by running the following command: time for i in {1..5}; do dd if=/mnt/home/greatfile$i.test of=/dev/null bs=4k; done The `real` parameter in the test output should be less than 15 seconds. #### Clear the Caches Unmount the NFS server to clear out any caches: umount /mnt/home ### Test Writing and Reading Small Files on the NFS Server Test small files as you did large files but with these sizes: - File size: 80 KB - Block size: 4 KB - Number of blocks: 20 Before starting the test, verify that the `/mnt/home` directory from the previous test is created. 1. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 2. Perform writes (five files of 80 KB each) on the NFS server: time for i in {1..30}; do dd if=/dev/zero of=/mnt/home/smallfile$i.test bs=4k count=20; done The `real` parameter in the test output should be around 2 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server: time for i in {1..30}; do dd if=/mnt/home/smallfile$i.test of=/dev/null bs=4k; done The important parameter in the test output is the `real` parameter, which should be around 2 seconds. #### Delete the Test Files When you finish the performance testing, delete the test files. 1. Remove all generated files by running: rm -f /mnt/home/\*.test 2. Unmount the NFS server: umount /mnt/home ## Override NFS Mount Options By default, Anypoint Platform PCE 4.2.0 sets the `noresvport` NFS mount option, which is strongly recommended for AWS and other cloud environments. If this default conflicts with your environment requirements, you can override the NFS mount options by providing a comma-separated string in the `platformConfiguration` section of your `input.yaml` file: ```yaml platformConfiguration: fileSystemBareMountOptionsOverride: "" fileSystemWcMountOptionsOverride: "" fileSystem01MountOptionsOverride: "" ``` For example, to remove the `noresvport` option, provide a string with your custom mount options: nfsvers=4.0,retrans=2,rsize=1048576,soft,sync,timeo=600,wsize=1048576 > [!NOTE] These fields are optional. If not specified, the default mount options are used. ## See Also - [Troubleshoot NFS](https://docs.mulesoft.com/private-cloud/4.2/troubleshoot-nfs) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.2/install-checklist) - [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Ops Center > For the full documentation index, see: https://docs.mulesoft.com/llms.txt The Ops Center is a web-based management tool that enables you to manage and monitor Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and perform the following types of tasks: - Add and remove servers - View logs and configure log forwarding - Monitor system stability and performance ## Access the Ops Center 1. Go to the following URL: https://{platform-host}:9500 2. Enter the admin user email address and password that you created during installation. ## Add Additional Servers Adding servers enables you to scale up your cluster, even after the installation is complete. 1. Log in to Ops Center. 2. Select Servers and then select Add Existing. 3. Select the type of node you want to add. The nodes available depend on the number of nodes in your installation. 4. Select Continue. Ops Center displays a command similar to the following: curl -s --tlsv1.2 --insecure "https://172.31.48.9:3009/t/8402020736a8a04dda29bc81dfcd5211/node" | sudo bash 5. Run the command that is displayed on the machine where you are installing the new server. - You must run this command from an account that has root access. - You can only perform this installation on one server at a time. After performing this command, the server is displayed in a list in Ops Center. 6. If necessary, specify the IP address and devices. 7. Select the server and then select Start Installation to begin installation of the new server. ## Remove a Server You can remove individual servers from the cluster where the platform is installed. > [!WARNING] Deleting a server also removes any applications that are deployed on the server. 1. Log in to Ops Center. 2. Select Servers. 3. Select the IP address of the server you want to delete and then select Delete. 4. Confirm that you want to delete the server. ## Configure Log Forwarding Anypoint Platform PCE enables you to forward server-level log data. Log forwarding is designed to forward your logs to a local third-party log solution. Forwarding log data to a cloud-based service like Splunk Cloud is not supported. > [!NOTE] This information is specific to forwarding log data from Anypoint Platform itself. For > instructions on how to forward log data from the Mule servers and applications that run on > Anypoint Platform, see > [Sending Data from Runtime Manager to External Analytics Software](https://docs.mulesoft.com/runtime-manager/sending-data-from-arm-to-external-analytics-software). Anypoint Platform uses `rsyslog` to handle logging. To forward logs to remote hosts, perform the following steps: 1. Log in to Ops Center. 2. In the upper right-hand corner, select the admin user name. ![ops center settings](/mulesoft/current/en-us/private-cloud/3.2/_images/ops-center-settings.png) 3. Select **Settings** and then select Logs. 4. Select **Create** or **New Connector** if you already have a log forwarder configured. 5. Configure your log forwarding as shown in the following example: kind: logforwarder version: v2 metadata: name: forwarder1 spec: address: 192.168.100.1:514 protocol: udp ## Monitor Stability and Performance Ops Center enables you to view data about the stability and performance of the platform. Metrics dashboards are available for the following application components: - Anypoint Services - Cluster - Object Store - Pods - PostgreSQL - Redis High-resolution metrics are stored for a default period of 30 days. Other metrics are retained from a period of 6 months up to 5 years. ### View Metrics 1. Log in to Ops Center. 2. Select the Monitoring tab. By default, Ops Center displays the Cluster dashboard. When viewing dashboards, navigate by selecting from a series of drop-down menus. The options that are available depend on the selected dashboard. For example, in the **Pods** dashboard, you can select a `namespace` and a `podname`. ### Change the Time Scale Displayed in the Charts To change the time scale displayed in the charts: 1. Select the currently selected time scale. 2. Select an option or provide start and end dates. ### Anypoint Services Dashboard The **Anypoint Services** dashboard displays metrics for the different platform sub-components, including Anypoint Exchange, API notebooks, or the Home page. You can filter data by a specific microservice. This dashboard displays the following metrics: - Processor - Memory - Network - Filesystem Each of these metrics is displayed by service. ### Cluster Information Dashboard The **Cluster Information** dashboard displays the following metrics for the various nodes that make up your platform cluster. You can filter data by `nodename`. - Overall CPU Usage - CPU Usage by Node - Individual CPU Usage - Memory Usage by Node - Individual Memory Usage - Overall Cluster Network Usage - Network Usage by Node - Individual Node Network Usage - Overall Cluster Filesystem Usage - Filesystem Usage by Node - Individual Node Filesystem Usage These metrics have different scopes, depending on the context: - Overall metrics show an aggregate number for the entire set of nodes. - Metrics that are _by node_ display each node as a separate curve on the same chart. - Individual node metrics are only displayed for the node that you selected in the **nodename** drop-down at the top of the section. ### Object Store Dashboard The **Object Store** dashboard displays data about the different nodes that make up the object store. It includes the following metrics: - Node Status - Read/Write Requests (in requests per second) - Read/Write Latency - Active Connections - Unavailable Exceptions - Disk Space Used Metrics that deal with Read/Write values display read and write values as separate curves. ### Pod Dashboard The **Pod** dashboard displays data for the individual Docker containers running different platform services. You can filter data by `namespace` or `podname`. The following metrics are displayed: - Individual CPU Usage - Individual Memory Usage - Individual Network Usage - File system Usage These metrics are displayed for the selected `namespace` and `podname`. ### PostgreSQL Dashboard The **PostgreSQL** dashboard displays data for the PostgreSQL server included in the platform. The following metrics are displayed: - Activity per type - Cache Hit Ratio - Active Connections - Buffers - Conflicts/Deadlocks - PostgreSQL Containers CPU Usage per Pod `Activity per type` displays different curves for rows fetched, returned, inserted, updated, and deleted, at the level of time granularity that you choose. ## Download Data To download the data displayed in this section as a JSON: 1. Select the gear icon at the top of the Ops Center. 2. Select **Export**. ![managing via the ops center f3b1c](/mulesoft/current/en-us/private-cloud/3.2/_images/managing-via-the-ops-center-f3b1c.png) Alternatively, you can select **Save as…​** to download this file with a custom name. You can also select the **View JSON** option to view this data without downloading it. ## Reset Your Password To change the password required to log into the Ops Center: 1. Enter the gravity utility: This utility is added during the platform installation. sudo gravity enter 2. Reset the password using the following command: gravity site --insecure reset-password This command returns the administrator email and password, for example: password for admin@example.com has been reset to: xxxxxxxxx ## See Also - [Add a Product License Using Ops Center](https://docs.mulesoft.com/private-cloud/3.2/ops-center-update-lic) - [Set up Alerts for CPU and Memory Usage](https://docs.mulesoft.com/private-cloud/3.2/config-alerts) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Preinstallation Verification > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Before installing or upgrading, your infrastructure team must review all prerequisite > information in the following list and verify that your environment meets the stated requirements. > If needed, contact your MuleSoft representative for assistance. - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-platform) - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Software Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-software) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) ## Verify Anypoint Platform Private Cloud Edition Nodes Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) on all nodes, run the `gravity check` script to verify that each node meets the minimum requirements for Anypoint Platform PCE, including CPU, RAM, and volume storage. 1. Copy the following information from the node where you will run the installer to each node of your cluster: - The `gravity` binary. - The `app.yaml` manifest used by the `gravity` binary. These files are located in the folder of the installer. 2. Run the following command from each node of your cluster: sudo ./gravity check --profile=app\_node app.yaml Valid values for `--profile` are: - app\_node - data\_node - general\_node (for 3 node installations) If the command returns no output, your environment is set up correctly. If any node fails to satisfy the minimum requirements, the command returns a list of failed checks and exits with a non-0 return code. --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Shared Filesystem Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Before you install Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), > your infrastructure team must review each of the following sections and verify that your > environment meets the stated requirements. If needed, contact your MuleSoft representative for > assistance to ensure your shared filesystem is configured correctly. Anypoint Platform PCE requires a `ReadWriteMany` (RWX) shared filesystem for the platform. Starting with Anypoint Platform PCE 4.3.0, you can provide this filesystem using either of the following storage modes: - **NFS mode** — Bind the platform to an existing NFS server through a pre-installed CSI driver. Use this mode when your organization already operates an NFS server (for example, AWS EFS, NetApp, Isilon, or a self-managed NFS server) that you want the platform to consume. - **StorageClass mode** — Provide a single RWX-capable Kubernetes `StorageClass` and let Anypoint Platform PCE dynamically provision the shared volumes. Use this mode when your cluster has an RWX-capable CSI driver such as CephFS available and you prefer not to operate an external NFS server. You choose the mode at installation time by setting the corresponding fields in the `platformConfiguration` section of your `input.yaml` file. The two modes are mutually exclusive. > [!IMPORTANT] The installer rejects changes to the storage mode on an existing cluster. Switching > between NFS and StorageClass mode after installation orphans the data stored on the previous > volumes. Choose the mode carefully before installing. ## CSI Driver Requirement (NFS Mode) When using NFS mode, a pre-installed Container Storage Interface (CSI) driver is a mandatory requirement. Anypoint Platform PCE uses the Kubernetes CSI driver to mount NFS volumes into pods, replacing the previous method that relied on host binaries. Some key benefits: - Enhanced support for environments where custom tools cannot be installed on host nodes. - Direct support for Amazon EKS users who can only use the Amazon EFS CSI Driver. > [!NOTE] The CSI driver requirement applies only to NFS mode. In StorageClass mode, the RWX-capable > `StorageClass` handles provisioning through its own CSI driver. ### Configure the CSI Driver In the `platformConfiguration` section of your `input.yaml` file, specify the CSI driver name: ```yaml platformConfiguration: fileSystemCsiDriverName: "" # CSI driver name (e.g., nfs.csi.k8s.io, efs.csi.aws.com) ``` ### EFS CSI Driver Configuration When using the Amazon EFS CSI Driver (for example, for Amazon EKS with EFS installations), provide four individual CSI volume handles. These volume handles must point to four access points that all target the same directory in the EFS: ```yaml platformConfiguration: fileSystemCsiDriverName: "efs.csi.aws.com" fileSystemWcCsiVolumeHandle: "" # CSI volume handle for wc filesystem fileSystem01CsiVolumeHandle: "" # CSI volume handle for 01 filesystem fileSystemBareCsiVolumeHandle: "" # CSI volume handle for bare filesystem fileSystemBackupRestoreCsiVolumeHandle: "" # CSI volume handle for backup-restore filesystem ``` Each volume handle value follows the format `::`, for example: `fs-0835b0d1eb7588eb1::fsap-03f6c68563ceb5acf`. > [!NOTE] When using the EFS CSI Driver, you do not need to provide `fileSystemDNS` or > `fileSystemPath` values. ## StorageClass Mode Starting with Anypoint Platform PCE 4.3.0, you can host the platform shared filesystem on any Kubernetes `StorageClass` that supports `ReadWriteMany` access mode, without operating an external NFS server. The `StorageClass` must be backed by an RWX-compliant CSI driver, for example CephFS (`ocs-storagecluster-cephfs`) on OpenShift Data Foundation. When you set `fileSystemStorageClass` in `platformConfiguration`, the installer provisions the platform’s shared volumes and the backup-restore volume as dynamically created `PersistentVolumeClaim` resources against the specified `StorageClass`. In this mode, the following NFS-mode fields are not used: - `fileSystemCsiDriverName` - `fileSystemDNS`, `fileSystemPath` - `fileSystemWcCsiVolumeHandle`, `fileSystem01CsiVolumeHandle`, `fileSystemBareCsiVolumeHandle`, `fileSystemBackupRestoreCsiVolumeHandle` ### Configure StorageClass Mode In the `platformConfiguration` section of your `input.yaml` file, specify the RWX-capable `StorageClass`: ```yaml platformConfiguration: fileSystemStorageClass: "" # e.g. ocs-storagecluster-cephfs ``` ### StorageClass Requirements The `StorageClass` you provide must: - Support the `ReadWriteMany` (RWX) access mode. - Be backed by an RWX-compliant CSI driver (for example, CephFS). - Be available and default-usable in the cluster where you install Anypoint Platform PCE. > [!IMPORTANT] For backup retention, use a `StorageClass` with `reclaimPolicy: Retain` for the > backup-restore volume. With `reclaimPolicy: Delete`, deleting the `PersistentVolumeClaim` also > destroys the underlying storage and any historical backup tarballs stored on it. ## Verify NFS Server Setup The following sections apply only when installing in NFS mode. If you are using StorageClass mode, skip to [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/install-workflow). Anypoint Platform PCE version 2.0 and later requires a Network File System (NFS) server and an available NFS endpoint. Your NFS server must meet the following requirements: - Must be running on a filesystem that has at least 250 GB available. - Must be using NFS 4.x. - Can be backed by a network-attached storage (NAS) system. - Must have port 2049 enabled for TCP. - Configure your network so that all of the PCE nodes can access the NFS server. The direct NFS client is Kubernetes 1.9 and requires conformance with NFS version 4. NFS 4.x does not use the `portmapper` service. ## Verify Required Port Is Open The NFS server requires port 2049 to be open for TCP. On each node, use netcat or a similar utility to verify that the required port is open. Netcat is a computer networking utility used to read and write to network connections using TCP. On each node, run: `nc -zv 2049` > [!NOTE] By default, the netcat tool is not included with Anypoint Platform PCE. If you receive a > `command not found` error, you can download and install netcat or a similar utility. When the responses from netcat are successful, you receive a response similar to the following example: \[ec2-user@ip-10-1-1-97 ~\]$ nc -zv 2049 Ncat: Version 7.50 ( https://nmap.org/ncat ) Ncat: Connected to 10.1.0.155:2049. Ncat: 0 bytes sent, 0 bytes received in 0.01 seconds. ## Verify NFS Server Is Mounted On each node, mount the NFS server in a temporary directory. 1. Switch to root: `sudo su` 2. To mount the NFS in a temporary directory, run the following commands and replace both the `` and the `` with your own values: `mkdir -p /mnt/home` `mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/` If the mount fails, you do not have the required permissions. ## Verify NFS Performance Perform careful analysis of your environment, both from the client and server point of view for optimal NFS performance. This involves: - Testing the performance and latency of your NFS server. Specifically, these operations will transfer data from the client (Anypoint Platform PCE environment) to the NFS server and measure how long each transfer takes. - Using the command `time dd if=/dev/zero of=/mnt/home/ bs= count=` to test NFS server performance while transferring files and specifying the following parameters: - Names `` - Block size and number of blocks in the transfer: `` and ``, respectively. - Validating how Anypoint Platform PCE handles NFS file transfers. Perform read and write operations with both large (128MB) and small (80KB) files. Perform all tests on each node. ### Test Writing and Reading Large Files on the NFS Server Perform the following read and write tests using large files: - File size: 128 MB - Block size: 4 KB - Number of blocks: 32768 While the tests are running, different files are generated (one file per test). 1. Create a directory in which to mount the NFS server: sudo su mkdir -p /mnt/home 2. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform writes (five files of 128 MB each) on the NFS server: time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done The command output is similar to the following: \[root@ip-10-1-1-97 ec2-user\]# time for i in {1..5}; do dd if=/dev/zero of=/mnt/home/greatfile$i.test bs=4k count=32768; done 32768+0 records in 32768+0 records out 134217728 bytes (134 MB) copied, 2.12805 s, 63.1 MB/s . . . real 0m8.378s user 0m0.034s sys 0m0.792s The `real` parameter in the test output should be less than 15 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount /mnt/home 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server by running the following command: time for i in {1..5}; do dd if=/mnt/home/greatfile$i.test of=/dev/null bs=4k; done The `real` parameter in the test output should be less than 15 seconds. #### Clear the Caches Unmount the NFS server to clear out any caches: umount /mnt/home ### Test Writing and Reading Small Files on the NFS Server Test small files as you did large files but with these sizes: - File size: 80 KB - Block size: 4 KB - Number of blocks: 20 Before starting the test, verify that the `/mnt/home` directory from the previous test is created. 1. Mount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 2. Perform writes (five files of 80 KB each) on the NFS server: time for i in {1..30}; do dd if=/dev/zero of=/mnt/home/smallfile$i.test bs=4k count=20; done The `real` parameter in the test output should be around 2 seconds. #### Validate Read Performance 1. Unmount the NFS server: umount 2. Remount the NFS server: mount -t nfs4 -o proto=tcp,port=2049 : /mnt/home/ 3. Perform reads from the NFS server: time for i in {1..30}; do dd if=/mnt/home/smallfile$i.test of=/dev/null bs=4k; done The important parameter in the test output is the `real` parameter, which should be around 2 seconds. #### Delete the Test Files When you finish the performance testing, delete the test files. 1. Remove all generated files by running: rm -f /mnt/home/\*.test 2. Unmount the NFS server: umount /mnt/home ## Override NFS Mount Options In NFS mode, Anypoint Platform PCE sets the `noresvport` NFS mount option by default, which is strongly recommended for AWS and other cloud environments. If this default conflicts with your environment requirements, you can override the NFS mount options by providing a comma-separated string in the `platformConfiguration` section of your `input.yaml` file: ```yaml platformConfiguration: fileSystemBareMountOptionsOverride: "" fileSystemWcMountOptionsOverride: "" fileSystem01MountOptionsOverride: "" ``` For example, to remove the `noresvport` option, provide a string with your custom mount options: nfsvers=4.0,retrans=2,rsize=1048576,soft,sync,timeo=600,wsize=1048576 > [!NOTE] These fields are optional. If not specified, the default mount options are used. ## See Also - [Troubleshoot NFS](https://docs.mulesoft.com/private-cloud/latest/troubleshoot-nfs) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/latest/install-checklist) - [Installing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Software Prerequisites > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To ensure the performance and stability of Anypoint Platform Private Cloud Edition (Anypoint Platform PCE), every node in your Anypoint Platform PCE environment must meet the software prerequisites described in this topic: - Operating system (OS) prerequisites - Third-party software prerequisites - MuleSoft software prerequisites > [!IMPORTANT] Before you install Anypoint Platform PCE, your infrastructure team must review each > of the following sections and verify that your environment meets the stated software requirements. > If needed, contact your MuleSoft representative for assistance. ## Operating System Prerequisites ### Verify Linux Distribution The following Linux distributions are supported: - Red Hat Enterprise Linux (RHEL) 7.8, 7.9, 8.1, 8.2, 8.3., 8.4, 8.5, 8.6, 8.7, 8.8, 9.2, 9.3, and 9.5. - CentOS 7.8 and 7.9 > [!NOTE] Anypoint Platform PCE no longer supports CentOS 8, as it has reached > [end-of-life](https://www.centos.org/centos-linux-eol/). ### Check for SELinux Custom Profiles Although SELinux is not required, Anypoint Platform PCE supports the default SELinux profile running in strict mode. If your environment uses a custom profile, it must provide an unconstrained user context when running the installer. Review any custom profiles with your MuleSoft representative. ### Enable and Load Kernel Modules You must enable the following kernel modules on each node: - `br_netfilter` - `ebtable_filter` - `ip_tables` - `iptable_filter` - `iptable_nat` - `overlay` For each kernel module, run a command similar to the following example to verify that each module is loaded. Replace `br_netfilter` with the name of the kernel module you are checking: root$ lsmod | grep br\_netfilter - If this command does not display any output, then the module is not loaded. In this case, load the module using `modprobe`: root$ modprobe br\_netfilter Ensure a module is loaded after a reboot by using the following command: root$ echo 'br\_netfilter' > /etc/modules-load.d/br\_netfilter.conf ### Verify Kernel Settings The following flags must be enabled: - `net.bridge.bridge-nf-call-iptables=1` - `net.ipv4.ip_forward=1` - `fs.may_detach_mounts=1` - `fs.inotify.max_queued_events=16384` - `fs.inotify.max_user_instances=8192` - `fs.inotify.max_user_watches=1048576` The kubelet requires the following settings, which do not update the Kernel settings automatically: - `vm.overcommit_memory=1` - `kernel.panic=10` - `kernel.panic_on_oops=1` Ensure that no processes on your system overrides any of these flags. To set a parameter, use `sysctl -w`, as shown in the following example: root$ sysctl -w net.ipv4.ip\_forward=1 To make the changes permanent for the next reboot, write the changes to a file under `/etc/sysctl.d`. Replace the setting in the following example with the setting you are configuring. echo net.ipv4.ip\_forward=1 >> /etc/sysctl.d/10-ipv4-forwarding.conf Permanently configure settings required by the kubelet: cat > /etc/sysctl.d/90-kubelet.conf << EOF vm.overcommit\_memory=1 kernel.panic=10 kernel.panic\_on\_oops=1 EOF Ensure that no processes or files conflict with those settings. ## Third Party Software Prerequisites ### Requirements When Firewalld Is Enabled Firewalld is an iptables controller that defines rules for persistent network traffic. If you are using firewalld with a Red Hat Enterprise Linux (RHEL) 7.3 operating system, you must enable forwarding on the `docker0` device. You must also forward any packets being sent from or to the 10.0.0.0/8 subnet. #### Determine If You Are Using Firewalld To determine if your system is using firewalld, run the following command on every node: systemctl status firewalld.service If firewalld is installed, the following output is displayed: $ sudo systemctl status firewalld.service \* firewalld.service - firewalld - dynamic firewall daemon Loaded: loaded (/usr/lib/systemd/system/firewalld.service; enabled; vendor preset: enabled) Active: inactive (dead) Docs: man:firewalld(1) If firewalld is not installed, an error message is displayed. #### Enable Forwarding on Docker Device To enable forwarding on the docker0 device, run the following commands: firewall-cmd --permanent --direct --add-rule ipv4 filter FORWARD 1 -o docker0 -j ACCEPT -m comment --comment "docker subnet" firewall-cmd --permanent --direct --add-rule ipv4 filter FORWARD 1 -s 10.0.0.0/8 -j ACCEPT -m comment --comment "docker subnet" To enable forwarding on the 10.0.0.0/8 subnet, run the following commands: firewall-cmd --permanent --direct --add-rule ipv4 filter INPUT 1 -s 10.0.0.0/8 -j ACCEPT -m comment --comment "docker subnet" firewall-cmd --permanent --direct --add-rule ipv4 filter INPUT 1 -d 10.0.0.0/8 -j ACCEPT -m comment --comment "docker subnet" ### Remove Local DNS Servers Remove local caching DNS servers listening on port 53, for example, `named`, `dnsmasq`, and `bind`. ### Install Required Tools - Tool to Verify NFS Server Ports are Open Netcat is recommended, but you can use your tool of choice to verify the correct ports are open on the NFS server. If you do not have `netcat`, you can install it from [https://netcat.sourceforge.net/download.php](https://netcat.sourceforge.net/download.php) or follow the instructions at [https://linoxide.com/linux-how-to/install-use-netcat-command-linux/](https://linoxide.com/linux-how-to/install-use-netcat-command-linux/). ### Remove Server Configuration Management Software Any Server Configuration Management software, such as Puppet, Chef, Ansible, and SaltStack, must be removed. This software usually changes kernel modules and settings that are required, and can also restart servers without following the correct shut down procedure, causing the system to enter a failed state. ## MuleSoft Software Prerequisites ### Obtain Install Archive and License Key Obtain the installer archive (`anypoint-3.2.x-installer.tar.gz`) and license key from your MuleSoft account representative. ### Verify IP Forwarding and Bridging To ensure proper cluster load balancing and routing, you must enable, correctly configure, and verify the following items: - Kernel IP forwarding To enable internal Kubernetes load balancing, you must enable IPv4 forwarding on all servers. - Bridge-netfilter Enable the Linux kernel to translate packets to and from hosted containers. - IP addresses for bridge-netfiltering are managed by the overlay network created by Kubernetes. Verify that the CIDR block used for that network is in the private IP address range. ## Verify SSL Certificate Requirements To use the Anypoint Platform PCE, you must provide SSL certificates. You can upload a certificate through the Anypoint Platform PCE UI. This certificate must be trusted by every machine that is connected to the platform. > [!WARNING] You must register the same SSL certificate on every node containing Mule runtime > engines managed by your installation. ### Uninstall Unsupported Software Some software causes conflicts with Anypoint Platform PCE. You must uninstall this software before performing Anypoint Platform PCE installation. #### Docker Uninstall Docker from all servers running the Anypoint Platform PCE. The Anypoint Platform PCE installation includes a Docker package officially supported by Kubernetes. ## See Also - [Environment Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-platform) - [Hardware Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-hardware) - [Network Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/prereq-network) - [NFS Server Prerequisites](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) - [Preinstallation Verification](https://docs.mulesoft.com/private-cloud/3.2/prereq-gravity-check) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Supported Configurations for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To ensure platform performance, stability, and high availability, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) supports two network configurations: 4-node and 7-node. To ensure high availability and performance, each node must run on its own server in both configurations. If you are using a virtual machine to host the nodes, you must ensure that each VM node is running on a separate physical server. Both the 4-node and 7-node configurations have two separate components: - Anypoint Platform Core - Monitoring Center Add-On (optional) ## 4-Node Configuration The 4-node configuration is the smallest configuration that Anypoint Platform PCE supports. This is the minimum number of nodes that can enable high-availability and failover. In this configuration, each node hosts the platform applications and services. You must configure a load balancer to use round-robin distribution of traffic among each of the four nodes. ![4-node Configuration](/mulesoft/current/en-us/private-cloud/3.2/_images/configuration-4-node.png) Three of the nodes host an instance of the database and object store. Hosting the database and object store enables persistence but requires larger minimum disk and memory requirements. Although three of the nodes contain a database, only one database is used as the primary. Applications on each node write to this database only. The other two database instances are hot standby instances of the primary database that take over as the primary database in case of failure. ## 7-Node Configuration In the 7-node configuration, four nodes are dedicated to hosting platform applications and services. The other three nodes host the database and object store instances. Although each node contains a database, only one of the database nodes is used as the primary. Applications on each node write to this database only. The other two database instances are hot standby instances of the primary database that take over as the primary database in case of failure. ![7-node Configuration](/mulesoft/current/en-us/private-cloud/3.2/_images/configuration-7-node.png) You must configure a load balancer to use round-robin distribution of traffic among all of the nodes in your cluster. ## Supported Platform Components and Configuration > [!WARNING] The following recommendations and limits are based on the following resources: > > - A 4-node environment consisting of 4 nodes of 8 cores and 32GB RAM, or > > - A 7-node environment consisting of 7 nodes of 8 cores and 32GB RAM. ### Mule Runtime Engine and Mule Applications The following table shows the recommended and maximum number of Mule instances and Mule applications that Anypoint Platform PCE supports:
4-node Configuration7-node Configuration

Maximum number of Mule instances per environment

500

500

Maximum number of Mule instances across all environments

1500

3000

Maximum number of Mule instances per server group

20

20

Maximum Mule instances per cluster

8

8

Mule applications under management per environment (recommended)

Under 200

Around 200

Mule applications under management per environment (maximum)

500

500

Mule instances under management (maximum)

500 single Mule runtime engine, or 250 Mule runtime clusters, consisting of 2 Mule runtime engines per cluster

1000 single Mule runtime engines, or 500 Mule runtime clusters, consisting of 2 Mule runtime engines per cluster

Mule applications per Mule runtime engine (recommended)

Fewer than 25

Fewer than 25

Mule applications per Mule runtime engine (maximum)

100

150

### APIs and API Manager

4-node Configuration

7-node Configuration

Maximum instances per API group

100

100

Client applications (per user)

5000

5000

Client applications (per root organization)

10000

10000

Maximum API instances manager per root organization

6000

12000

Applied policies

12000

24000

Created contracts

12000

24000

For both 4-node and 7-node configurations, the following apply: - Maximum instances per API group - Applications - A single API implementation with one policy counts as 0.5 APIs. - A single clustered proxy with up to 5 policies counts as 1. - All Mule instances are polling every 60 seconds. ### Access Management
4-node Configuration7-node Configuration

Maximum number of supported business groups

75

N/A

Maximum number of users

10000

N/A

See the Access Management documentation for more information on [identity management and supported providers](https://docs.mulesoft.com/access-management/external-identity). ### Exchange
4-node Configuration7-node Configuration

Maximum Exchange assets

60,000

60,000

Maximum file size when publishing asset through UI

5 MB

5 MB

Maximum uncompressed file size

50 MB

50 MB

Maximum file size when publishing asset through Maven

200 MB

200

Maximum number of dependencies for a single asset

100

100

Maximum icon size

2 MB

2 MB

Maximum number of resources (attachments) per asset portal

100

100

Maximum resource size for attachments to the asset portal

2 MB

2 MB

Maximum indexable asset content for RAML or asset portal

1 MB

1 MB

Maximum number of categories per organization

500

500

Maximum number of custom configurations per organization

500

500

Maximum number of pages in a portal

50

50

Maximum length of page name

128 characters

128 characters

Maximum review title length

30 characters

30 characters

Maximum review text length

2048 characters

2048 characters

The 1 MB maximum indexable asset content for RAML or asset portal means that if your portal has more than 1MB in text, only the first 1MB will be indexed, and if your portal has more than 1 MB in RAML specifications, only the first 1 MB will be indexed. ### API Designer and API Mocking Service The following apply to both 4-node and 7-node configurations.

Maximum number of projects

Limited by version control system

Maximum number of concurrent users

16

Maximum number of mocking service instances

One per API version

Maximum number of mocking service users

16

### Anypoint Monitoring Anypoint Monitoring is an optional, add-on component. See [Anypoint Monitoring on Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/anypoint-monitoring). The following limitations apply when using Anypoint Monitoring with Anypoint Platform PCE. #### Enhanced API Analytics Monitoring Limits - API Summary Metrics - Retention: 1 year - API IP Address Metrics - Number of IP addresses per API per minute: limited to 20 unique IP addresses. - Retention: 1 hour - API Endpoint Metrics - Number of endpoints per API per minute: limited to 100 unique endpoints - Retention: 1 hour #### Application Monitoring Limits - Number of endpoints per application per minute: 5 unique endpoints - Number of client IP addresses per application per minute: 5 unique endpoints #### Custom Metrics - Up to 10000 alerts - System performs #### Advanced Alert Limits - Up to 20 alerts #### Performance Limits - Number of Mule instances: 3000 - Number of applications: 12000 - Number of application replicas: up to 12000. An application can run on multiple Mule instances in a server group or cluster. - Number of APIs: up to 120000 - Number of custom metrics: 50 #### Examples of Supported Parameters - 6000 applications, each with 1 extra replica corresponds to 12000 application replicas - 2000 applications, each with 6 extra replicas corresponds to 12000 application replicas - 12000 applications, with no replicas corresponds to 12000 application replicas #### Hardware Requirements - 3 Nodes - CPU: 32 cores - Memory: 128 GB - Disk requirements: - 4TB InfluxDB volume for each AMV node - 500 GB Logstash volume per AMV node - 500 GB insight-cassandra-meld volume per AMV node - IOPS: 1500 IPS per volume (1 voolume each for Influx, logstash, and insight-cassandra-meld) - Network: 10 GB/second #### License Requirements - The Influx Enterprise license (included in the Anypoint Monitoring subscription) is required to enable monitoring. ### Anypoint Visualizer Anypoint Platform PCE does not the support the following Anypoint Visualizer features: - List of applications If an application does not use Anypoint Monitoring, Anypoint Visualizer cannot track it on Anypoint Platform PCE. - Policy visualization ## See Also - [Anypoint Monitoring](https://docs.mulesoft.com/private-cloud/3.2/anypoint-monitoring) - [Anypoint Visualizer](https://docs.mulesoft.com/visualizer/) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Supported Configurations for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To ensure platform performance, stability, and high availability, Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) supports two network configurations in Kubernetes (K8s) nodes: 4-node and 7-node. ## Anypoint Platform PCE 4.0 Architecture The diagram illustrates MuleSoft’s planes (control plane and runtime plane) and their general traffic flow within the Anypoint Platform PCE context: ![](/mulesoft/current/en-us/private-cloud/4.0/_images/pce4-diagram-architecture.png) Control plane: - A subset of Anypoint services is deployed in Kubernetes (customer’s choice - Anypoint Platform PCE v 4.0 supports RKE2, EKS-A, OpenShift). - This constitutes the MuleSoft Control Plane, a web-based management and administration tool for the lifecycle of MuleSoft integration apps and APIs. - Starting with Anypoint Platform PCE v4.0, use a dedicated Kubernetes control plane to separate cluster management components from customer workloads. Runtime plane: - MuleSoft offers various runtime planes - Within Anypoint Platform PCE context, the test uses the Hybrid runtime plane to deploy and manage runtimes via the MuleSoft control plane. The following recommendations and limits are based on these resources: \* A small (4-node) environment consisting of 4 K8s worker nodes of 8 cores and 32GB RAM and 3 dedicated K8s control plane nodes, or \* A large (7-node) environment consisting of 7 K8s worker nodes of 8 cores and 32GB RAM and 3 dedicated K8s control plane nodes ## Mule Runtime Engine and Mule Applications The following table shows the recommended and maximum number of Mule instances and Mule applications that Anypoint Platform PCE supports:
DescriptionSmall cluster 4-node ConfigurationLarge cluster 7-node Configuration

Maximum number of Mule instances per environment

500

500

Maximum number of Mule instances across all environments

1500

3000

Maximum number of Mule instances per server group

20

20

Maximum Mule instances per cluster

8

8

Mule applications under management per environment (recommended)

Under 200

Around 200

Mule applications under management per environment (maximum)

500

500

Mule applications per Mule runtime engine (recommended)

Fewer than 25

Fewer than 25

Mule applications per Mule runtime engine (maximum)

100

150

Number of Mule applications

6000

12000

Number of Mule application replicas. An application can run on multiple Mule instances in a server group or cluster.

12000

24000

## APIs and API Manager The following table show the maximum instances per API, number of client applications, applied policies and created contracts that Anypoint Platform PCE supports:

Description

Small cluster 4-node Configuration

Large cluster 7-node Configuration

Maximum instances per API group

100

100

Client applications (per user)

5000

5000

Client applications (per root organization)

10000

10000

Maximum API instances manager per root organization

6000

12000

Applied policies

18000

36000

Created contracts

12000

24000

## Access Management The following table show the maximum number of business groups and number of users of Access Management that Anypoint Platform PCE supports:
DescriptionSmall cluster 4-node ConfigurationLarge cluster 7-node Configuration

Maximum number of supported business groups

75

N/A

Maximum number of users

10000

N/A

See the Access Management documentation for more information on [identity management and supported providers](https://docs.mulesoft.com/access-management/external-identity). ## Exchange The following table show the maximum number of exchange assets and file size information that Anypoint Platform PCE supports:
DescriptionSmall cluster 4-node ConfigurationLarge cluster 7-node Configuration

Maximum Exchange assets

60,000

60,000

Maximum file size when publishing asset through UI

5 MB

5 MB

Maximum uncompressed file size

50 MB

50 MB

Maximum file size when publishing asset through Maven

200 MB

200

Maximum number of dependencies for a single asset

100

100

Maximum icon size

2 MB

2 MB

Maximum number of resources (attachments) per asset portal

100

100

Maximum resource size for attachments to the asset portal

2 MB

2 MB

Maximum indexable asset content for RAML or asset portal

1 MB

1 MB

Maximum number of categories per organization

500

500

Maximum number of custom configurations per organization

500

500

Maximum number of pages in a portal

50

50

Maximum length of page name

128 characters

128 characters

Maximum review title length

30 characters

30 characters

Maximum review text length

2048 characters

2048 characters

## API Designer and API Mocking Service The following apply to both 4-node and 7-node configurations.

Maximum number of projects

Limited by version control system

Maximum number of concurrent users

16

Maximum number of mocking service instances

One per API version

Maximum number of mocking service users

16

## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.0/) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Supported Configurations for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt These are the supported configurations for Anypoint Platform Private Cloud (Anypoint Platform PCE). ## Anypoint Platform PCE 4.1 Architecture The diagram illustrates MuleSoft’s planes (control plane and runtime plane) and their general traffic flow within the Anypoint Platform PCE context: ![Anypoint Platform PCE 4.1 architecture](/mulesoft/current/en-us/private-cloud/4.1/_images/pce4-diagram-architecture.png) Control plane: - A subset of Anypoint Platform services is deployed to a supported Kubernetes distribution (RKE2, EKS-A, or OpenShift). - This constitutes the MuleSoft Control Plane, a web-based management and administration tool for the lifecycle of MuleSoft integration apps and APIs. - Starting with Anypoint Platform PCE v4.0, use a dedicated Kubernetes control plane to separate cluster management components from customer workloads. Runtime plane: - MuleSoft offers various runtime planes. - Within the Anypoint Platform PCE context, use the Hybrid runtime plane to deploy and manage runtimes via the MuleSoft control plane. ## Mule Runtime Engine and Mule Applications This table shows the recommended and maximum number of Mule instances and Mule applications that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum number of Mule instances per environment

500

Maximum number of Mule instances across all environments

4000

Maximum number of Mule instances per server group

20

Maximum Mule instances per cluster

8

Mule applications under management per environment (recommended)

Fewer than 200

Mule applications under management per environment (maximum)

500

Mule applications per Mule runtime engine (recommended)

Fewer than 25

Mule applications per Mule runtime engine (maximum)

100

Number of Mule applications

15000

Number of Mule application replicas. An application can run on multiple Mule instances in a server group or cluster.

30000

## APIs and API Manager This table shows the maximum number of instances per API, number of client applications, applied policies, and created contracts that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum instances per API group

100

Client applications (per user)

5000

Client applications (per root organization)

10000

Maximum API instances managed per root organization

12000

Applied policies

36000

Created contracts

24000

## Access Management This table shows the maximum number of business groups and number of Access Management users that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum number of supported business groups

75

Maximum number of users

10000

Maximum number of client providers

25

## Exchange This table shows the maximum number of Anypoint Exchange (Exchange) assets and file size information that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum Exchange assets

60,000

Maximum file size when publishing asset through UI

5 MB

Maximum uncompressed file size

50 MB

Maximum file size when publishing asset through Maven

200 MB

Maximum number of dependencies for a single asset

100

Maximum icon size

2 MB

Maximum number of resources (attachments) per asset portal

100

Maximum resource size for attachments to the asset portal

2 MB

Maximum indexable asset content for RAML or asset portal

1 MB

Maximum number of categories per organization

500

Maximum number of custom configurations per organization

500

Maximum number of pages in a portal

50

Maximum length of page name

128 characters

Maximum review title length

30 characters

Maximum review text length

2048 characters

## API Designer and API Mocking Service

Maximum number of projects

Limited by version control system

Maximum number of concurrent users

16

Maximum number of mocking service instances

One per API version

Maximum number of mocking service users

16

## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.1/) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Supported Configurations for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt These are the supported configurations for Anypoint Platform Private Cloud (Anypoint Platform PCE). ## Anypoint Platform PCE 4.2 Architecture The diagram illustrates MuleSoft’s planes (control plane and runtime plane) and their general traffic flow within the Anypoint Platform PCE context: ![Anypoint Platform PCE 4.2 architecture](/mulesoft/current/en-us/private-cloud/4.2/_images/pce4-diagram-architecture.png) Control plane: - A subset of Anypoint Platform services is deployed to a supported Kubernetes distribution (RKE2, EKS-A, or OpenShift). - This constitutes the MuleSoft Control Plane, a web-based management and administration tool for the lifecycle of MuleSoft integration apps and APIs. - Starting with Anypoint Platform PCE v4.0, use a dedicated Kubernetes control plane to separate cluster management components from customer workloads. Runtime plane: - MuleSoft offers various runtime planes. - Within the Anypoint Platform PCE context, use the Hybrid runtime plane to deploy and manage runtimes via the MuleSoft control plane. ## Mule Runtime Engine and Mule Applications This table shows the recommended and maximum number of Mule instances and Mule applications that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum number of Mule instances per environment

500

Maximum number of Mule instances across all environments

4000

Maximum number of Mule instances per server group

20

Maximum Mule instances per cluster

8

Mule applications under management per environment (recommended)

Fewer than 200

Mule applications under management per environment (maximum)

500

Mule applications per Mule runtime engine (recommended)

Fewer than 25

Mule applications per Mule runtime engine (maximum)

100

Number of Mule applications

15000

Number of Mule application replicas. An application can run on multiple Mule instances in a server group or cluster.

30000

## APIs and API Manager This table shows the maximum number of instances per API, number of client applications, applied policies, and created contracts that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum instances per API group

100

Client applications (per user)

5000

Client applications (per root organization)

10000

Maximum API instances managed per root organization

12000

Applied policies

36000

Created contracts

24000

## Access Management This table shows the maximum number of business groups and number of Access Management users that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum number of supported business groups

75

Maximum number of users

10000

Maximum number of client providers

25

## Exchange This table shows the maximum number of Anypoint Exchange (Exchange) assets and file size information that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum Exchange assets

60,000

Maximum file size when publishing asset through UI

5 MB

Maximum uncompressed file size

50 MB

Maximum file size when publishing asset through Maven

200 MB

Maximum number of dependencies for a single asset

100

Maximum icon size

2 MB

Maximum number of resources (attachments) per asset portal

100

Maximum resource size for attachments to the asset portal

2 MB

Maximum indexable asset content for RAML or asset portal

1 MB

Maximum number of categories per organization

500

Maximum number of custom configurations per organization

500

Maximum number of pages in a portal

50

Maximum length of page name

128 characters

Maximum review title length

30 characters

Maximum review text length

2048 characters

## API Designer and API Mocking Service

Maximum number of projects

Limited by version control system

Maximum number of concurrent users

16

Maximum number of mocking service instances

One per API version

Maximum number of mocking service users

16

## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.2/) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Supported Configurations for Anypoint Platform PCE > For the full documentation index, see: https://docs.mulesoft.com/llms.txt These are the supported configurations for Anypoint Platform Private Cloud (Anypoint Platform PCE). ## Anypoint Platform PCE 4.3 Architecture The diagram illustrates MuleSoft’s planes (control plane and runtime plane) and their general traffic flow within the Anypoint Platform PCE context: ![Anypoint Platform PCE 4.3 architecture, with the load balancer, Kubernetes worker and control-plane nodes, Mule runtimes, and external storage, SMTP, registry, NFS, and identity provider](/mulesoft/current/en-us/private-cloud/latest/_images/pce4.3-architecture-diagram.png) Control plane: - A subset of Anypoint Platform services is deployed to a supported Kubernetes distribution (RKE2, EKS-A, or OpenShift). - This constitutes the MuleSoft Control Plane, a web-based management and administration tool for the lifecycle of MuleSoft integration apps and APIs. - Starting with Anypoint Platform PCE v4.0, use a dedicated Kubernetes control plane to separate cluster management components from customer workloads. Runtime plane: - MuleSoft offers various runtime planes. - Within the Anypoint Platform PCE context, use the Hybrid runtime plane to deploy and manage runtimes via the MuleSoft control plane. ## Mule Runtime Engine and Mule Applications This table shows the recommended and maximum number of Mule instances and Mule applications that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum number of Mule instances per environment

500

Maximum number of Mule instances across all environments

4000

Maximum number of Mule instances per server group

20

Maximum Mule instances per cluster

8

Mule applications under management per environment (recommended)

Fewer than 200

Mule applications under management per environment (maximum)

500

Mule applications per Mule runtime engine (recommended)

Fewer than 25

Mule applications per Mule runtime engine (maximum)

100

Number of Mule applications

15000

Number of Mule application replicas. An application can run on multiple Mule instances in a server group or cluster.

30000

## APIs and API Manager This table shows the maximum number of instances per API, number of client applications, applied policies, and created contracts that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum instances per API group

100

Client applications (per user)

5000

Client applications (per root organization)

10000

Maximum API instances managed per root organization

12000

Applied policies

36000

Created contracts

24000

## Access Management This table shows the maximum number of business groups and number of Access Management users that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum number of supported business groups

75

Maximum number of users

10000

Maximum number of client providers

25

## Exchange This table shows the maximum number of Anypoint Exchange (Exchange) assets and file size information that Anypoint Platform PCE supports:
DescriptionSupported Capacity

Maximum Exchange assets

60,000

Maximum file size when publishing asset through UI

5 MB

Maximum uncompressed file size

50 MB

Maximum file size when publishing asset through Maven

200 MB

Maximum number of dependencies for a single asset

100

Maximum icon size

2 MB

Maximum number of resources (attachments) per asset portal

100

Maximum resource size for attachments to the asset portal

2 MB

Maximum indexable asset content for RAML or asset portal

1 MB

Maximum number of categories per organization

500

Maximum number of custom configurations per organization

500

Maximum number of pages in a portal

50

Maximum length of page name

128 characters

Maximum review title length

30 characters

Maximum review text length

2048 characters

## API Designer and API Mocking Service

Maximum number of projects

Limited by version control system

Maximum number of concurrent users

16

Maximum number of mocking service instances

One per API version

Maximum number of mocking service users

16

## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/latest/) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Troubleshoot Mule Runtime Engine Registration > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you encounter a Mule runtime engine registration error when [creating a server in Runtime Manager](https://docs.mulesoft.com/runtime-manager/servers-create), you might need to configure Mule to trust the TLS certificates you are using. To do so: 1. Get the public certificate for Anypoint Platform Private Cloud Edition. 2. Add the certificate to the truststore of the JVM you are using for the Mule runtime engine. Using the following example, set `$ANYPOINT_DNS` to your DNS for Anypoint Platform Private Cloud Edition. Replace the keystore location and password for your JVM truststore. $ keytool -printcert -sslserver $ANYPOINT\_DNS:443 -rfc > /tmp/PCE\_Certificate.cer $ keytool -importcert -alias anypoint.my-company -keystore -storepass changeit -file /tmp/PCE\_Certificate.cer For more information about the CA certificate truststore, refer to your JVM vendor’s documentation: for example, [Oracle documentation](https://docs.oracle.com/). --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Troubleshoot NFS Errors > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Setting up and configuring an NFS server should be performed by your infrastructure > team. If you experience problems with any of the following instructions, contact your > infrastructure team for assistance. You must have a connection from each node to NFS. ## “Server Not Responding” Use a tool such as ping, traceroute, or tracepath to verify that the client and server machines can communicate with each other. If communication fails between the client and server machines, examine the network interface card (NIC) settings using either `ifconfig` or `ethtool` to verify the IP settings. ## “Access Denied” or "Permission Denied" Check the export permissions for the NFS file system. You can do this from the client or from the server. From the client: `showmount -e server_name` From the server: `exportfs -a` ## "rpc mount export: RPC: Timed out" The error message `Unable to access file system at [NFS SERVER]: rpc mount export: RPC: Timed out` is caused by a DNS name resolution issue. NFS (RPC) needs reverse name resolution. If the NFS server or client cannot resolve their name, this error occurs. Check the DNS configuration and the `/etc/hosts` configuration. ## Mounting the NFS Server Fails If the mount fails, you are probably missing some NFS-required libraries. If this happens, install them by running: `yum install nfs-utils nfs-utils-lib` ### "mount failed: Not Permitted" The error message `mount failed: exit status 32. Mounting command …​ not permitted` might occur when trying to mount using the `noresvport` option because the ports might be limited. If this happens, remove the `noresvport` option and add the `insecure` flag to the mount command. ## See Also - [Verify NFS Server Communication](https://docs.mulesoft.com/private-cloud/3.2/verify-nfs) - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Troubleshooting Mule Runtime Engine Registration > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you encounter a Mule runtime engine (Mule) registration error when [creating a server in Runtime Manager](https://docs.mulesoft.com/runtime-manager/servers-create), configure Mule to trust the TLS certificates you are using. To do so: 1. Get the public certificate for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). 2. Add the certificate to the truststore of the JVM you are using for Mule. Using the following example, set `$ANYPOINT_DNS` to your DNS for Anypoint Platform PCE. Replace the keystore location and password for your JVM truststore. $ keytool -printcert -sslserver $ANYPOINT\_DNS:443 -rfc > /tmp/PCE\_Certificate.cer $ keytool -importcert -alias anypoint.my-company -keystore -storepass changeit -file /tmp/PCE\_Certificate.cer For more information about the CA certificate truststore, see your JVM vendor’s documentation: [Oracle documentation](https://docs.oracle.com/). ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Troubleshooting Mule Runtime Engine Registration > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you encounter a Mule runtime engine (Mule) registration error when [creating a server in Runtime Manager](https://docs.mulesoft.com/runtime-manager/servers-create), configure Mule to trust the TLS certificates you are using. To do so: 1. Get the public certificate for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). 2. Add the certificate to the truststore of the JVM you are using for Mule. Using the following example, set `$ANYPOINT_DNS` to your DNS for Anypoint Platform PCE. Replace the keystore location and password for your JVM truststore. $ keytool -printcert -sslserver $ANYPOINT\_DNS:443 -rfc > /tmp/PCE\_Certificate.cer $ keytool -importcert -alias anypoint.my-company -keystore -storepass changeit -file /tmp/PCE\_Certificate.cer For more information about the CA certificate truststore, see your JVM vendor’s documentation: [Oracle documentation](https://docs.oracle.com/). ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Troubleshooting Mule Runtime Engine Registration > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you encounter a Mule runtime engine (Mule) registration error when [creating a server in Runtime Manager](https://docs.mulesoft.com/runtime-manager/servers-create), configure Mule to trust the TLS certificates you are using. To do so: 1. Get the public certificate for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). 2. Add the certificate to the truststore of the JVM you are using for Mule. Using the following example, set `$ANYPOINT_DNS` to your DNS for Anypoint Platform PCE. Replace the keystore location and password for your JVM truststore. $ keytool -printcert -sslserver $ANYPOINT\_DNS:443 -rfc > /tmp/PCE\_Certificate.cer $ keytool -importcert -alias anypoint.my-company -keystore -storepass changeit -file /tmp/PCE\_Certificate.cer For more information about the CA certificate truststore, see your JVM vendor’s documentation: [Oracle documentation](https://docs.oracle.com/). ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Troubleshooting Mule Runtime Engine Registration > For the full documentation index, see: https://docs.mulesoft.com/llms.txt If you encounter a Mule runtime engine (Mule) registration error when [creating a server in Runtime Manager](https://docs.mulesoft.com/runtime-manager/servers-create), configure Mule to trust the TLS certificates you are using. To do so: 1. Get the public certificate for Anypoint Platform Private Cloud Edition (Anypoint Platform PCE). 2. Add the certificate to the truststore of the JVM you are using for Mule. Using the following example, set `$ANYPOINT_DNS` to your DNS for Anypoint Platform PCE. Replace the keystore location and password for your JVM truststore. $ keytool -printcert -sslserver $ANYPOINT\_DNS:443 -rfc > /tmp/PCE\_Certificate.cer $ keytool -importcert -alias anypoint.my-company -keystore -storepass changeit -file /tmp/PCE\_Certificate.cer For more information about the CA certificate truststore, see your JVM vendor’s documentation: [Oracle documentation](https://docs.oracle.com/). ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Troubleshooting NFS Errors > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Setting up and configuring an NFS server should be performed by your infrastructure > team. If you experience problems with any of the following instructions, contact your > infrastructure team for assistance. You must have a connection from each node to NFS. ## “Server Not Responding” Use a tool such as ping, traceroute, or tracepath to verify that the client and server machines can communicate with each other. If communication fails between the client and server machines, examine the network interface card (NIC) settings using either `ifconfig` or `ethtool` to verify the IP settings. ## “Access Denied” or "Permission Denied" Check the export permissions for the NFS file system. You can do this from the client or from the server. From the client: `showmount -e server_name` From the server: `exportfs -a` ## "rpc mount export: RPC: Timed out" The error message `Unable to access file system at [NFS SERVER]: rpc mount export: RPC: Timed out` is caused by a DNS name resolution issue. NFS (RPC) needs reverse name resolution. If the NFS server or client cannot resolve their name, this error occurs. Check the DNS configuration and the `/etc/hosts` configuration. ## Mounting the NFS Server Fails If the mount fails, you are probably missing some NFS-required libraries. If this happens, install them by running: `yum install nfs-utils nfs-utils-lib` ### "mount failed: Not Permitted" The error message `mount failed: exit status 32. Mounting command …​ not permitted` might occur when trying to mount using the `noresvport` option because the ports might be limited. If this happens, remove the `noresvport` option and add the `insecure` flag to the mount command. ## See Also - [Verify NFS Server Communication](https://docs.mulesoft.com/private-cloud/4.0/verify-nfs) - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Troubleshooting NFS Errors > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Setting up and configuring an NFS server should be performed by your infrastructure > team. If you experience problems with any of the following instructions, contact your > infrastructure team for assistance. You must have a connection from each node to NFS. ## “Server Not Responding” Use a tool such as ping, traceroute, or tracepath to verify that the client and server machines can communicate with each other. If communication fails between the client and server machines, examine the network interface card (NIC) settings using either `ifconfig` or `ethtool` to verify the IP settings. ## “Access Denied” or "Permission Denied" Check the export permissions for the NFS file system. You can do this from the client or from the server. From the client: `showmount -e server_name` From the server: `exportfs -a` ## "rpc mount export: RPC: Timed out" The error message `Unable to access file system at [NFS SERVER]: rpc mount export: RPC: Timed out` is caused by a DNS name resolution issue. NFS (RPC) needs reverse name resolution. If the NFS server or client cannot resolve their name, this error occurs. Check the DNS configuration and the `/etc/hosts` configuration. ## Mounting the NFS Server Fails If the mount fails, you are probably missing some NFS-required libraries. If this happens, install them by running: `yum install nfs-utils nfs-utils-lib` ### "mount failed: Not Permitted" The error message `mount failed: exit status 32. Mounting command …​ not permitted` might occur when trying to mount using the `noresvport` option because the ports might be limited. If this happens, remove the `noresvport` option and add the `insecure` flag to the mount command. ## See Also - [Verify NFS Server Communication](https://docs.mulesoft.com/private-cloud/4.1/verify-nfs) - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Troubleshooting NFS Errors > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Setting up and configuring an NFS server should be performed by your infrastructure > team. If you experience problems with any of the following instructions, contact your > infrastructure team for assistance. You must have a connection from each node to NFS. ## “Server Not Responding” Use a tool such as ping, traceroute, or tracepath to verify that the client and server machines can communicate with each other. If communication fails between the client and server machines, examine the network interface card (NIC) settings using either `ifconfig` or `ethtool` to verify the IP settings. ## “Access Denied” or "Permission Denied" Check the export permissions for the NFS file system. You can do this from the client or from the server. From the client: `showmount -e server_name` From the server: `exportfs -a` ## "rpc mount export: RPC: Timed out" The error message `Unable to access file system at [NFS SERVER]: rpc mount export: RPC: Timed out` is caused by a DNS name resolution issue. NFS (RPC) needs reverse name resolution. If the NFS server or client cannot resolve their name, this error occurs. Check the DNS configuration and the `/etc/hosts` configuration. ## Mounting the NFS Server Fails If the mount fails, you are probably missing some NFS-required libraries. If this happens, install them by running: `yum install nfs-utils nfs-utils-lib` ### "mount failed: Not Permitted" The error message `mount failed: exit status 32. Mounting command …​ not permitted` might occur when trying to mount using the `noresvport` option because the ports might be limited. If this happens, remove the `noresvport` option and add the `insecure` flag to the mount command. ## See Also - [Verify NFS Server Communication](https://docs.mulesoft.com/private-cloud/4.2/verify-nfs) - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Troubleshooting NFS Errors > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!IMPORTANT] Setting up and configuring an NFS server should be performed by your infrastructure > team. If you experience problems with any of the following instructions, contact your > infrastructure team for assistance. > [!NOTE] This page applies only when installing Anypoint Platform PCE in NFS mode. If you are using > StorageClass mode (Anypoint Platform PCE 4.3.0 and later), troubleshoot storage issues through > your `StorageClass` provider and CSI driver instead. See > [StorageClass Mode](https://docs.mulesoft.com/private-cloud/latest/verify-nfs#storage-class-mode). You must have a connection from each node to NFS. ## “Server Not Responding” Use a tool such as ping, traceroute, or tracepath to verify that the client and server machines can communicate with each other. If communication fails between the client and server machines, examine the network interface card (NIC) settings using either `ifconfig` or `ethtool` to verify the IP settings. ## “Access Denied” or "Permission Denied" Check the export permissions for the NFS file system. You can do this from the client or from the server. From the client: `showmount -e server_name` From the server: `exportfs -a` ## "rpc mount export: RPC: Timed out" The error message `Unable to access file system at [NFS SERVER]: rpc mount export: RPC: Timed out` is caused by a DNS name resolution issue. NFS (RPC) needs reverse name resolution. If the NFS server or client cannot resolve their name, this error occurs. Check the DNS configuration and the `/etc/hosts` configuration. ## Mounting the NFS Server Fails If the mount fails, you are probably missing some NFS-required libraries. If this happens, install them by running: `yum install nfs-utils nfs-utils-lib` ### "mount failed: Not Permitted" The error message `mount failed: exit status 32. Mounting command …​ not permitted` might occur when trying to mount using the `noresvport` option because the ports might be limited. If this happens, remove the `noresvport` option and add the `insecure` flag to the mount command. ## See Also - [Verify NFS Server Communication](https://docs.mulesoft.com/private-cloud/latest/verify-nfs) - [Known Issues](https://help.salesforce.com/s/issues) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Update InfluxDB License > For the full documentation index, see: https://docs.mulesoft.com/llms.txt To update the InfluxDB license, contact MuleSoft Professional Services for the new license and the update instructions. --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Update Nodes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] Where possible, we changed noninclusive terms to align with our company value of Equality. > We maintained certain terms to avoid any effect on customer implementations. ## Before You Begin - Set up and test the backup procedure on all clusters. - Perform a backup before attempting node maintenance to preserve data in the event of a failure. ## Cluster and Application Updates > [!IMPORTANT] You must perform the following steps on one Anypoint Platform Private Cloud Edition > node at a time to minimize or eliminate component breakage during the update process. You must > complete the entire process on each node before the next node is addressed. After each node is > updated, verify the health status of all cluster components and applications before proceeding > with the next node. ### Verify Cluster Health Status Verify that the cluster is healthy before starting the update process. Otherwise, many of the assumptions made in this process are not valid. 1. Verify the output from the following commands: ```none gravity status etcdctl cluster-health ``` If the output from the previous commands mentions a `degraded` state or `errors`, fix all issues before proceeding. 2. Verify that all pods are running: ```none kubectl get pod -o wide --all-namespaces | grep -vi Running ``` If the output from the previous command lists any pod not in "Running" state, then fix all issues before proceeding. ### Install Updates Select a node on which to perform the following steps: 1. Retrieve the name of the target node: ```none kubectl get nodes ``` 2. Drain the node: ```none kubectl drain --delete-local-data --ignore-daemonsets ``` 3. After the node is drained, identify the gravity unit name: ```none sudo systemctl list-units --plain | awk '/planet-master/{print $1}' ``` The previous command outputs a `systemd` unit name, which is used as the `$unit_name` value in the following step. 4. Stop the gravity unit: ```none systemctl stop $unit_name systemctl status $unit_name ``` 5. After the `status` subcommand indicates that the gravity unit is stopped, install the updates and reboot the server. ### Perform Post-update Verification 1. Verify that all systems correctly restarted: ```none systemctl status --failed ``` The previous command should return zero units, which means that your system restart was successful and all units started. If any units failed to start, fix all issues before proceeding. 2. Check the gravity `systemd` units: ```none systemctl list-units '*gravity__gravitational*' ``` 3. Check for errors in the logs by running `journalctl`. 4. If everything started correctly, run `gravity enter` and verify the results. 5. Check the cluster status again: ```none gravity status etcdctl cluster-health gravity get pod -o wide--all-namespaces | grep -vi Running ``` ### Rejoin the Node 1. Check the gravity Planet service and teleport status: ```none systemctl list-units '*gravity__gravitational*' ``` 2. If the status of the gravity service is not `loaded`, `active`, or `running`, then you must enable it: ```none systemctl enable --now ``` 3. Uncordon the node to set the target node to again schedule pods. This can take a few minutes to complete. ```none kubectl uncordon ``` 4. Verify cluster status: ```none gravity status ``` This command should display a `healthy` status for all nodes of the cluster. After you verify that the node has rejoined the cluster and is in healthy status, repeat the process for the next node in the cluster until all nodes are updated. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Updating Nodes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] Where possible, we changed noninclusive terms to align with our company value of Equality. > We maintained certain terms to avoid any effect on customer implementations. ## Before You Begin - Set up and test the backup procedure on all clusters. - Perform a backup before attempting node maintenance to preserve data in the event of a failure. ## Update Nodes for Clusters See the following topics for details on each provider: - [Rancher Kubernetes Engine (RKE2) OS Upgrade Process](https://support.scc.suse.com/s/redirect?language=en_US&id=000021301) - OpenShift: Starting with version 4.16, the Machine Config Operator manages the node’s OS updates as part of the cluster update process. For more on node updates and the Machine Config Operator, see: - [Openshift Updates](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html) - [Openshift Update Service](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html#update-service-about_understanding-openshift-updates) - [Machine Config Operator Functionality](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/how-updates-work.html) ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Updating Nodes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] Where possible, we changed noninclusive terms to align with our company value of Equality. > We maintained certain terms to avoid any effect on customer implementations. ## Before You Begin - Set up and test the backup procedure on all clusters. - Perform a backup before attempting node maintenance to preserve data in the event of a failure. ## Update Nodes for Clusters See the following topics for details on each provider: - [Rancher Kubernetes Engine (RKE2) OS Upgrade Process](https://support.scc.suse.com/s/redirect?language=en_US&id=000021301) - OpenShift: Starting with version 4.16, the Machine Config Operator manages the node’s OS updates as part of the cluster update process. For more on node updates and the Machine Config Operator, see: - [Openshift Updates](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html) - [Openshift Update Service](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html#update-service-about_understanding-openshift-updates) - [Machine Config Operator Functionality](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/how-updates-work.html) ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Updating Nodes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] Where possible, we changed noninclusive terms to align with our company value of Equality. > We maintained certain terms to avoid any effect on customer implementations. ## Before You Begin - Set up and test the backup procedure on all clusters. - Perform a backup before attempting node maintenance to preserve data in the event of a failure. ## Update Nodes for Clusters See the following topics for details on each provider: - [Rancher Kubernetes Engine (RKE2) OS Upgrade Process](https://support.scc.suse.com/s/redirect?language=en_US&id=000021301) - OpenShift: Starting with version 4.16, the Machine Config Operator manages the node’s OS updates as part of the cluster update process. For more on node updates and the Machine Config Operator, see: - [Openshift Updates](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html) - [Openshift Update Service](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html#update-service-about_understanding-openshift-updates) - [Machine Config Operator Functionality](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/how-updates-work.html) ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Updating Nodes > For the full documentation index, see: https://docs.mulesoft.com/llms.txt > [!NOTE] Where possible, we changed noninclusive terms to align with our company value of Equality. > We maintained certain terms to avoid any effect on customer implementations. ## Before You Begin - Set up and test the backup procedure on all clusters. - Perform a backup before attempting node maintenance to preserve data in the event of a failure. ## Update Nodes for Clusters See the following topics for details on each provider: - [Rancher Kubernetes Engine (RKE2) OS Upgrade Process](https://support.scc.suse.com/s/redirect?language=en_US&id=000021301) - OpenShift: Starting with version 4.16, the Machine Config Operator manages the node’s OS updates as part of the cluster update process. For more on node updates and the Machine Config Operator, see: - [Openshift Updates](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html) - [Openshift Update Service](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/intro-to-updates.html#update-service-about_understanding-openshift-updates) - [Machine Config Operator Functionality](https://docs.openshift.com/en/container-platform/4.16/updating/understanding_updates/how-updates-work.html) ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Upgrade Workflow > For the full documentation index, see: https://docs.mulesoft.com/llms.txt ## MuleSoft Professional Services Required Contact MuleSoft professional services prior to upgrading. MuleSoft professional services must be involved throughout the upgrade process. Before an Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) upgrade is performed, you must verify that your working installation meets any additional requirements of the new Anypoint Platform PCE version, such as opening additional ports or adding a new port for the load balancer. > [!NOTE] When upgrading from a previous Anypoint Platform PCE version, all policies available in > the previous version remain. Policies are not replaced, but additional policies might be included > with the upgrade. Refer to [About Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/3.2/) for additional information about policies included with Anypoint Platform PCE. Refer to the [Anypoint Platform Private Cloud Edition Release Notes](https://docs.mulesoft.com/release-notes/pce/anypoint-private-cloud-release-notes) for additional information about the most recent Anypoint Platform PCE version. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Upgrading Anypoint Platform PCE v3.2.3 to v4.0.0 > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before you upgrade Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) to version 4.0, verify that your working installation meets any additional requirements of the new Anypoint Platform PCE version. Note that if your Anypoint Platform PCE version is earlier than version 3.2.3, you must first upgrade to version 3.2.3 and then upgrade to version 4.0. See the following topics: - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.0/) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.0/install-checklist) > [!NOTE] Contact MuleSoft professional services prior to upgrading. MuleSoft professional services > must be involved throughout the upgrade process. ## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.0/) - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.0/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Upgrading Anypoint Platform PCE v3.2.5 to v4.1.x > For the full documentation index, see: https://docs.mulesoft.com/llms.txt Before you upgrade Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) to version 4.1, verify that your working installation meets any additional requirements of the new Anypoint Platform PCE version. If your current PCE version is earlier than 3.2.4, upgrade to 3.2.5 to minimize potential issues. See the following topics: - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.1/) - [Anypoint Platform PCE Prerequisites](https://docs.mulesoft.com/private-cloud/4.1/install-checklist) > [!NOTE] Contact MuleSoft professional services prior to upgrading. MuleSoft professional services > must be involved throughout the upgrade process. ## See Also - [Anypoint Platform PCE Overview](https://docs.mulesoft.com/private-cloud/4.1/) - [Installing Anypoint Platform Private Cloud Edition (Anypoint Platform PCE)](https://docs.mulesoft.com/private-cloud/4.1/install-workflow) --- product: Anypoint Platform Private Cloud Edition version: 3.2 --- # Upload Custom Policies and Publish Assets to Exchange Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt There are differences between Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and the cloud version of Anypoint Platform that you must consider when uploading custom policies and publishing assets to Anypoint Exchange using the Maven client. See the following documentation for the cloud version of Anypoint Platform for details: - [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) Perform the following steps for Anypoint Platform PCE: 1. Add the server and repository sections in the `settings.xml` file to set up the archetype for your project: The following snippet shows the relevant sections: ```xml ... apce-repository myusername mypassword ... apce-repository Anypoint PCE Repository https://${ANYPOINT_DNS}/exchange/maven ... ``` In the following examples, ${ANYPOINT\_DNS} is set to your DNS for Anypoint Platform PCE. The `username` and `password` must correspond with your installation. 2. Generate the archetype: Follow the standard process: [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started). For example: mvn -Parchetype-repository archetype:generate \\ -DarchetypeGroupId=org.mule.tools \\ -DarchetypeArtifactId=api-gateway-custom-policy-archetype \\ -DarchetypeVersion=1.0.0 \\ -DgroupId=${orgId} \\ -DartifactId=${policyName} \\ -Dversion=1.0.0-SNAPSHOT \\ -Dpackage=mule-policy 3. Replace the `exchange.url` in the `pom.xml` file. By default, `exchange.url` points to the cloud version of Anypoint Platform. `maven.anypoint.mulesoft.com` must point to the Anypoint Platform PCE endpoint: https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven The `repositoryId` must match the one you configured for Maven in the `settings.xml` file. 4. Package the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Packaging a Custom Policy](https://docs.mulesoft.com/mule-gateway/policies-custom-package). 5. Upload the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Uploading a Custom Policy to Exchange](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange). If you receive an error similar to: \[ERROR\] Failed to execute goal org.apache.maven.plugins:maven-deploy-plugin:2.8.2:deploy (default-deploy) on project ${policyName}: Failed to retrieve remote metadata ${groupId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml: Could not transfer metadata ${orgId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml from/to exchange-server (https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven): PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target -> \[Help 1\] This means the certificate of the platform is not trusted by the local Maven. The solution is to trust the certificate in the environment where Maven is running, as shown in the following example: keytool -printcert -sslserver ${ANYPOINT\_DNS}:443 -rfc > /tmp/anypoint.crt keytool -importcert -alias my-company.com -keystore /etc/ssl/certs/java/cacerts -storepass changeit -file /tmp/anypoint.crt In the previous example, $ANYPOINT\_DNS is set to the DNS for Anypoint Platform PCE. Replace the location and password for your JVM’s truststore. 6. Verify your policy was uploaded by directly visiting `https://${ANYPOINT_DNS}/exchange/?type=policy` or `https://${ANYPOINT_DNS}/exchange/${orgID}/${artifactId}/`. For an API managed by Mule 4 or later, the policy is displayed in the **Apply Policy** dialog. --- product: Anypoint Platform Private Cloud Edition version: 4.0 --- # Uploading Custom Policies and Publish Assets to Exchange Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt There are differences between Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and the cloud version of Anypoint Platform that you must consider when uploading custom policies and publishing assets to Anypoint Exchange using the Maven client. For details of the cloud version of Anypoint Platform, see: - [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) Perform the following steps for Anypoint Platform PCE: 1. Add the server and repository sections in the `settings.xml` file to set up the archetype for your project: The following snippet shows the relevant sections: ```xml ... apce-repository myusername mypassword ... apce-repository Anypoint PCE Repository https://${ANYPOINT_DNS}/exchange/maven ... ``` In the following examples, ${ANYPOINT\_DNS} is set to your DNS for Anypoint Platform PCE. The `username` and `password` must correspond with your installation. 2. Generate the archetype: Follow the standard process: [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started). For example: mvn -Parchetype-repository archetype:generate \\ -DarchetypeGroupId=org.mule.tools \\ -DarchetypeArtifactId=api-gateway-custom-policy-archetype \\ -DarchetypeVersion=1.0.0 \\ -DgroupId=${orgId} \\ -DartifactId=${policyName} \\ -Dversion=1.0.0-SNAPSHOT \\ -Dpackage=mule-policy 3. Replace the `exchange.url` in the `pom.xml` file. By default, `exchange.url` points to the cloud version of Anypoint Platform. `maven.anypoint.mulesoft.com` must point to the Anypoint Platform PCE endpoint: https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven The `repositoryId` must match the one you configured for Maven in the `settings.xml` file. 4. Package the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Packaging a Custom Policy](https://docs.mulesoft.com/mule-gateway/policies-custom-package). 5. Upload the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Uploading a Custom Policy to Exchange](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange). If you receive an error similar to: \[ERROR\] Failed to execute goal org.apache.maven.plugins:maven-deploy-plugin:2.8.2:deploy (default-deploy) on project ${policyName}: Failed to retrieve remote metadata ${groupId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml: Could not transfer metadata ${orgId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml from/to exchange-server (https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven): PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target -> \[Help 1\] This means the certificate of the platform is not trusted by the local Maven. The solution is to trust the certificate in the environment where Maven is running, as shown in the following example: keytool -printcert -sslserver ${ANYPOINT\_DNS}:443 -rfc > /tmp/anypoint.crt keytool -importcert -alias my-company.com -keystore /etc/ssl/certs/java/cacerts -storepass changeit -file /tmp/anypoint.crt In the previous example, $ANYPOINT\_DNS is set to the DNS for Anypoint Platform PCE. Replace the location and password for your JVM’s truststore. 6. Verify your policy was uploaded by directly visiting `https://${ANYPOINT_DNS}/exchange/?type=policy` or `https://${ANYPOINT_DNS}/exchange/${orgID}/${artifactId}/`. For an API managed by Mule 4 or later, the policy is displayed in the **Apply Policy** dialog. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.0/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.1 --- # Uploading Custom Policies and Publish Assets to Exchange Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt There are differences between Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and the cloud version of Anypoint Platform that you must consider when uploading custom policies and publishing assets to Anypoint Exchange using the Maven client. For details of the cloud version of Anypoint Platform, see: - [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) Perform the following steps for Anypoint Platform PCE: 1. Add the server and repository sections in the `settings.xml` file to set up the archetype for your project: The following snippet shows the relevant sections: ```xml ... apce-repository myusername mypassword ... apce-repository Anypoint PCE Repository https://${ANYPOINT_DNS}/exchange/maven ... ``` In the following examples, ${ANYPOINT\_DNS} is set to your DNS for Anypoint Platform PCE. The `username` and `password` must correspond with your installation. 2. Generate the archetype: Follow the standard process: [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started). For example: mvn -Parchetype-repository archetype:generate \\ -DarchetypeGroupId=org.mule.tools \\ -DarchetypeArtifactId=api-gateway-custom-policy-archetype \\ -DarchetypeVersion=1.0.0 \\ -DgroupId=${orgId} \\ -DartifactId=${policyName} \\ -Dversion=1.0.0-SNAPSHOT \\ -Dpackage=mule-policy 3. Replace the `exchange.url` in the `pom.xml` file. By default, `exchange.url` points to the cloud version of Anypoint Platform. `maven.anypoint.mulesoft.com` must point to the Anypoint Platform PCE endpoint: https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven The `repositoryId` must match the one you configured for Maven in the `settings.xml` file. 4. Package the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Packaging a Custom Policy](https://docs.mulesoft.com/mule-gateway/policies-custom-package). 5. Upload the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Uploading a Custom Policy to Exchange](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange). If you receive an error similar to: \[ERROR\] Failed to execute goal org.apache.maven.plugins:maven-deploy-plugin:2.8.2:deploy (default-deploy) on project ${policyName}: Failed to retrieve remote metadata ${groupId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml: Could not transfer metadata ${orgId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml from/to exchange-server (https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven): PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target -> \[Help 1\] This means the certificate of the platform is not trusted by the local Maven. The solution is to trust the certificate in the environment where Maven is running, as shown in the following example: keytool -printcert -sslserver ${ANYPOINT\_DNS}:443 -rfc > /tmp/anypoint.crt keytool -importcert -alias my-company.com -keystore /etc/ssl/certs/java/cacerts -storepass changeit -file /tmp/anypoint.crt In the previous example, $ANYPOINT\_DNS is set to the DNS for Anypoint Platform PCE. Replace the location and password for your JVM’s truststore. 6. Verify your policy was uploaded by directly visiting `https://${ANYPOINT_DNS}/exchange/?type=policy` or `https://${ANYPOINT_DNS}/exchange/${orgID}/${artifactId}/`. For an API managed by Mule 4 or later, the policy is displayed in the **Apply Policy** dialog. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.1/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.2 --- # Uploading Custom Policies and Publish Assets to Exchange Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt There are differences between Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and the cloud version of Anypoint Platform that you must consider when uploading custom policies and publishing assets to Anypoint Exchange using the Maven client. For details of the cloud version of Anypoint Platform, see: - [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) Perform the following steps for Anypoint Platform PCE: 1. Add the server and repository sections in the `settings.xml` file to set up the archetype for your project: The following snippet shows the relevant sections: ```xml ... apce-repository myusername mypassword ... apce-repository Anypoint PCE Repository https://${ANYPOINT_DNS}/exchange/maven ... ``` In the following examples, ${ANYPOINT\_DNS} is set to your DNS for Anypoint Platform PCE. The `username` and `password` must correspond with your installation. 2. Generate the archetype: Follow the standard process: [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started). For example: mvn -Parchetype-repository archetype:generate \\ -DarchetypeGroupId=org.mule.tools \\ -DarchetypeArtifactId=api-gateway-custom-policy-archetype \\ -DarchetypeVersion=1.0.0 \\ -DgroupId=${orgId} \\ -DartifactId=${policyName} \\ -Dversion=1.0.0-SNAPSHOT \\ -Dpackage=mule-policy 3. Replace the `exchange.url` in the `pom.xml` file. By default, `exchange.url` points to the cloud version of Anypoint Platform. `maven.anypoint.mulesoft.com` must point to the Anypoint Platform PCE endpoint: https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven The `repositoryId` must match the one you configured for Maven in the `settings.xml` file. 4. Package the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Packaging a Custom Policy](https://docs.mulesoft.com/mule-gateway/policies-custom-package). 5. Upload the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Uploading a Custom Policy to Exchange](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange). If you receive an error similar to: \[ERROR\] Failed to execute goal org.apache.maven.plugins:maven-deploy-plugin:2.8.2:deploy (default-deploy) on project ${policyName}: Failed to retrieve remote metadata ${groupId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml: Could not transfer metadata ${orgId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml from/to exchange-server (https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven): PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target -> \[Help 1\] This means the certificate of the platform is not trusted by the local Maven. The solution is to trust the certificate in the environment where Maven is running, as shown in the following example: keytool -printcert -sslserver ${ANYPOINT\_DNS}:443 -rfc > /tmp/anypoint.crt keytool -importcert -alias my-company.com -keystore /etc/ssl/certs/java/cacerts -storepass changeit -file /tmp/anypoint.crt In the previous example, $ANYPOINT\_DNS is set to the DNS for Anypoint Platform PCE. Replace the location and password for your JVM’s truststore. 6. Verify your policy was uploaded by directly visiting `https://${ANYPOINT_DNS}/exchange/?type=policy` or `https://${ANYPOINT_DNS}/exchange/${orgID}/${artifactId}/`. For an API managed by Mule 4 or later, the policy is displayed in the **Apply Policy** dialog. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/4.2/operating-about) --- product: Anypoint Platform Private Cloud Edition version: 4.3 is-latest-version: true --- # Uploading Custom Policies and Publish Assets to Exchange Using the Maven Client > For the full documentation index, see: https://docs.mulesoft.com/llms.txt There are differences between Anypoint Platform Private Cloud Edition (Anypoint Platform PCE) and the cloud version of Anypoint Platform that you must consider when uploading custom policies and publishing assets to Anypoint Exchange using the Maven client. For details of the cloud version of Anypoint Platform, see: - [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started) - [Uploading Custom Policies](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange) - [Maven facade API](https://anypoint.mulesoft.com/exchange/portals/anypoint-platform/f1e97bc6-315a-4490-82a7-23abe036327a.anypoint-platform/exchange-maven-facade-api-http/) Perform the following steps for Anypoint Platform PCE: 1. Add the server and repository sections in the `settings.xml` file to set up the archetype for your project: The following snippet shows the relevant sections: ```xml ... apce-repository myusername mypassword ... apce-repository Anypoint PCE Repository https://${ANYPOINT_DNS}/exchange/maven ... ``` In the following examples, ${ANYPOINT\_DNS} is set to your DNS for Anypoint Platform PCE. The `username` and `password` must correspond with your installation. 2. Generate the archetype: Follow the standard process: [Getting started with Custom Policies development](https://docs.mulesoft.com/mule-gateway/policies-custom-getting-started). For example: mvn -Parchetype-repository archetype:generate \\ -DarchetypeGroupId=org.mule.tools \\ -DarchetypeArtifactId=api-gateway-custom-policy-archetype \\ -DarchetypeVersion=1.0.0 \\ -DgroupId=${orgId} \\ -DartifactId=${policyName} \\ -Dversion=1.0.0-SNAPSHOT \\ -Dpackage=mule-policy 3. Replace the `exchange.url` in the `pom.xml` file. By default, `exchange.url` points to the cloud version of Anypoint Platform. `maven.anypoint.mulesoft.com` must point to the Anypoint Platform PCE endpoint: https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven The `repositoryId` must match the one you configured for Maven in the `settings.xml` file. 4. Package the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Packaging a Custom Policy](https://docs.mulesoft.com/mule-gateway/policies-custom-package). 5. Upload the custom policy. Follow the same process as documented for the cloud version of Anypoint Platform in [Uploading a Custom Policy to Exchange](https://docs.mulesoft.com/mule-gateway/policies-custom-upload-to-exchange). If you receive an error similar to: \[ERROR\] Failed to execute goal org.apache.maven.plugins:maven-deploy-plugin:2.8.2:deploy (default-deploy) on project ${policyName}: Failed to retrieve remote metadata ${groupId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml: Could not transfer metadata ${orgId}:${policyName}:1.0.0-SNAPSHOT/maven-metadata.xml from/to exchange-server (https://${ANYPOINT\_DNS}/exchange/maven/api/v1/organizations/${orgId}/maven): PKIX path building failed: sun.security.provider.certpath.SunCertPathBuilderException: unable to find valid certification path to requested target -> \[Help 1\] This means the certificate of the platform is not trusted by the local Maven. The solution is to trust the certificate in the environment where Maven is running, as shown in the following example: keytool -printcert -sslserver ${ANYPOINT\_DNS}:443 -rfc > /tmp/anypoint.crt keytool -importcert -alias my-company.com -keystore /etc/ssl/certs/java/cacerts -storepass changeit -file /tmp/anypoint.crt In the previous example, $ANYPOINT\_DNS is set to the DNS for Anypoint Platform PCE. Replace the location and password for your JVM’s truststore. 6. Verify your policy was uploaded by directly visiting `https://${ANYPOINT_DNS}/exchange/?type=policy` or `https://${ANYPOINT_DNS}/exchange/${orgID}/${artifactId}/`. For an API managed by Mule 4 or later, the policy is displayed in the **Apply Policy** dialog. ## See Also - [Managing Anypoint Platform PCE](https://docs.mulesoft.com/private-cloud/latest/operating-about)