Requeue a Job or Task

Requeuing a job or task inserts it back into the job queue. You can requeue, Canceled, and, in some cases, Failed jobs or tasks.

You can use the following procedures to requeue jobs or tasks using HPC Cluster Manager.

Note

To complete this procedure, you must be the job owner or a cluster administrator.

To requeue a job

  1. In Job Management, in the Navigation Pane, click All Jobs.

  2. In the job list, select one or more jobs.

  3. Right-click your selection and click Requeue Job.

To requeue a task

  1. In Job Management, in the Navigation Pane, click All Jobs.

  2. In the job list, click a job. The list of tasks for that job appears in the Detail Pane.

  3. In the task list, select one or more tasks.

  4. Right-click your selection and click Requeue Task.

Guidelines for requeuing jobs and tasks

  • Requeuing a job requeues tasks that are Queued or Failed.

  • When you requeue a task, the job that contains the task must be Running. Otherwise, the job itself must be requeued in order to rerun the task.

  • Finished jobs and tasks cannot be requeued. You can make a copy of a Finished job or task to run it again. For more information, see Copy a Job or Task.

Additional considerations

  • By default, the HPC Job Scheduler Service automatically requeues a job or task that fails because of a system failure, such as a node reboot. If you do not want tasks within the job to be requeued automatically, set the task property Rerunnable to false.

  • For information about using HPC Cluster Manager, see Overview of HPC Cluster Manager.

Additional references