To ensure that the most up-to-date data is available by a particular time or to run a job during non-business hours, schedule a batch data transform.
- In Data Transforms, from Recently Viewed, click
next to the data transform
name, and select Schedule. - Under Run Mode, click Scheduled.
- Select the frequency and time interval.
- Day Intervals—When you schedule a transform to run at an interval of days, the
schedule is based on the first day of the month.
For example, on April 19, you schedule a batch data transform to run every 10 days. From the first day of the month, 10 days ahead is April 11, and 10 days after the 11th is April 21, the first date to occur after the schedule was set on April 19.
The transform runs on these dates: 4/21, 5/1, 5/11, 5/21, and 5/31. On 6/1, the schedule resets on the first day of the month and continues running every 10 days.
- Hourly Intervals—When you schedule a transform to run at an interval of hours,
the schedule begins at the start time specified and runs until the end of the day.
For example, at 8 PM, you schedule a batch data transform to run every 4 hours, beginning at 6 PM on each day. The transform starts running on the day that the user creates the schedule at the first interval after 6 PM. The transform runs at these times: 10 PM, 6 PM, and 10 PM.
Even though you created the schedule at 8 PM, the batch data transform runs 4 hours after the defined start time of 6 PM, which is 10 PM. The next day, it runs at 6 PM, followed by a second run at 10 PM. Because the schedule completes at the end of the day, the transform doesn’t run again until the next day at 6 PM.
- Day Intervals—When you schedule a transform to run at an interval of days, the
schedule is based on the first day of the month.
- Select the start time and time zone.
If a scheduled run becomes due while the previous run of the same batch data transform is still in progress, that run is skipped, runs never overlap, and the next run proceeds as scheduled.The scheduled time only determines when a run is queued, not when it starts; depending on concurrency limits and available capacity, the run may actually start slightly after the scheduled time.
