--- schema_version: "1" id: f9d3e0dbd643d531156784678cac4e35a3e6e0a3fda3a9808a76a3fd29f9b19b title: Hardware collection: tableau version: current locale: en-gb portal_url: https://docs.salesforce.com/tableau/current/en-gb/server/rmt-incidents-hardware.md canonical_url: https://help.tableau.com/current/server/en-gb/rmt-incidents-hardware.htm source_site: help.tableau.com content_type: documentation content_hash: 0a21b817b359a0aea6b0002856d198dbdb12f80ef8ec963471a65b3b0b2f7616 description: Configure hardware incidents to monitor server CPU, memory, and disk in RMT. --- [Skip to main content](#main-content) # Hardware Incidents _Hardware_ incidents monitor the server itself. These can be used to help identify server issues that may affect Tableau Server’s performance. You can set thresholds for the following issues: - CPU Usage - Available Memory - Memory Usage - Free Disk Space The following incidents are configured by default when you install a new Tableau Resource Monitoring Tool: - When the available disk space falls below 10 GB for 10 minutes or more a warning incident is logged, and when available disk falls below 5 GB for 10 minutes or more, a critical incident is logged. - When available memory falls below 8 GB for over 10 minutes, a warning incident is logged. - When the CPU usage for the entire server is 80% or more for 5 minutes, a warning incident is logged. **Note:** Memory related incidents are configured in binary multiples of bytes. You can configure thresholds using the RMT Server web interface or by updating the configuration file `config.json`. To set the thresholds for hardware incidents, under the **Admin** menu, select **Configuration**, and go to the **Incidents** tab. For **CPU Usage**, set the following: To set the thresholds for hardware incidents, under the **Admin** menu, select **Configuration**, and go to the **Incidents** tab.

Key	Required?	Description
Severity	Required	See Incident Severity Level.
Process	Required	The threshold applies to the entire Tableau Server or for a single process as specified.
Start Threshold	Required	The CPU usage must surpass the value specified before an incident is created and monitored. Set the percent and the duration for this threshold.
End Threshold	Optional	The CPU usage that must fall below the value specified before an incident is considered resolved.

For **Available Memory**, set the following:

Key	Required?	Description
Severity	Required	See Incident Severity Level.
Start Threshold	Required	The available memory must fall below the value specified before an incident is created and monitored. Set the percent and the duration for this threshold.
End Threshold	Optional	The available memory must be above the value specified before an incident is considered resolved.

For **Memory Usage**, set the following:

Key	Required?	Description
Severity	Required	See Incident Severity Level.
Process	Required	The threshold applies to the entire Tableau Server or for a single process as specified.
Start Threshold	Required	The memory usage must be equal to the value specified before an incident is created and monitored. Set the percent and the duration for this threshold.
End Threshold	Optional	The memory usage must be below the value specified before an incident is considered resolved.

For **Free Disk Space**, set the following:

Key	Required?	Description
Severity	Required	See Incident Severity Level.
Start Threshold	Required	The free disk space must fall below the value specified before an incident is created and monitored. Set the percent and the duration for this threshold.
End Threshold	Optional	The free disk space must be above the value specified before an incident is considered resolved.

For **Disk Queue Length**, set the following:

Key	Required?	Description
Severity	Required	See Incident Severity Level.
Start Threshold	Required	The disk queue length must be equal to the value specified before an incident is created and monitored. Set the percent and the duration for this threshold.
End Threshold	Optional	The disk queue length must be below the value specified before an incident is considered resolved.

An example `config.json` snipped defining two hardware incidents: ``` { "monitoring": { "incidents": { "triggers": [ { "counter": "DiskSpaceAvailableKB", "severity": "warning", "threshold": 1048576 }, { "counter": "ProcessorTimePercent", "severity": "warning", "threshold": 0.95, "thresholdDuration": 300000, "endThreshold": 0.90, "endThresholdDuration": 5000 } ] } } } ``` - The **DiskSpaceAvailableKB** incident will trigger a warning once the available disk space falls below 10 GB. - The **ProcessorTimePercent** incident will trigger a warning once the CPU has had at least 95% utilisation for over 5 minutes. The incident will be considered resolved once the CPU is below 90% utilisation for 5 seconds. The default settings may or may not meet your requirements and can be changed based on your environment. As an example, for an environment whose identifier was “staging-environment” to trigger a warning when the available disk space falls below 2 GB, the configuration would look like: ``` { "environments": { "staging-environment": { "monitoring": { "incidents": { "triggers": [ { "counter": "DiskSpaceAvailableKB", "severity": "warning", "threshold": 2097152 } ] } } } } } ```

Key	Data Type	Required?	Description
`counter`	String	Required	The identifier for the hardware incident to monitor. available options are: `ProcessorTimePercent` `DiskSpaceAvailableKB` `DiskQueueTotalLength` `MemoryAvailableKB` `MemoryCommittedKB`
`severity`	String	Optional	See Incident Severity Level. Default value: `Warning`
`threshold`	Number	Required	The threshold that must be surpassed before an incident is monitored.
`thresholdDuration`	Number	Optional	The amount of time in milliseconds to monitor the situation before triggering an incident. If not specified, an incident will be triggered as soon as the `threshold` is reached.
`endThreshold`	Number	Optional	The threshold that must be surpassed before an incident is considered resolved.
endThresholdDuration	Number	Optional	The amount of time in milliseconds to monitor the situation before completing the incident. If not specified, an incident will be resolved as soon as the `endThreshold` is reached. If `endThreshold` is not defined, then `threshold` is used.

## Who can do this Resource Monitoring Tool Administrator or a Resource Monitoring Tool user with **Server/Environment Management** role.