How to restart a Hyperscale job
A Hyperscale Job Execution can be restarted if the Load or Post Load process of the Execution has failed or been cancelled.
To restart a Hyperscale Execution, use the: /executions/{id}/restart PUT endpoint of JobExecution API.
When restarting a Hyperscale Execution, if:
-
The Load process of the Execution has FAILED or has been CANCELLED, the Job Execution restart endpoint will trigger a restart of the Load process (this will also include a restart of the Pre-Load process).
-
The Load process of the Execution has already SUCCEEDED but the Post Load process has FAILED or has been CANCELLED, the Job Execution restart endpoint will trigger a restart of Post Load process only.
To restart a Hyperscale Job Execution, the following conditions on the Job Execution should be met:
-
The status of the Execution should be either FAILED or CANCELLED.
-
The Unload and Masking process of the Execution should have SUCCEEDED status.
-
The cleanup of the Execution should not have been triggered.
-
The retain execution data flag (retain_execution_data) of the Job should not be set to NO.
-
Hyperscale containers should not be restarted after the failure or cancellation of the Job Execution.