In the development console of fully managed Flink, you can suspend or terminate a job that is in the RUNNING state. This topic describes how to suspend and terminate a job.

Background information

  • Suspend a job: This operation does not clear the job status. For example, if the job that you suspend is running a COUNT operation, the COUNT operation continues from the last successful checkpoint after you resume the job. If you change the resource configuration, the configuration takes effect after the job is suspended and then resumed.
  • Terminate a job: This operation clears the job status. For example, if the job that you want to terminate is running a COUNT operation, the COUNT operation starts from 0 after you terminate the job and then restart it. If you modify the SQL logic, change the job version, add or delete parameters to or from the WITH clause, or add parameters for a job, the changes take effect after you terminate the job and then restart it.

Suspend a job

  1. Log on to the Realtime Compute for Apache Flink console.
  2. On the Fully Managed Flink tab, find the desired workspace and click Development Console in the Actions column.
  3. In the left-side navigation pane, click Deployments.
  4. Find the job that you want to suspend, and click Suspend in the Actions column.

Terminate a job

  1. Log on to the Realtime Compute for Apache Flink console.
  2. On the Fully Managed Flink tab, find the desired workspace and click Development Console in the Actions column.
  3. In the left-side navigation pane, click Deployments.
  4. Find the job that you want to terminate, and click Cancel in the Actions column.