Creating and aborting jobs

Migrate for Compute Engine defines a job as an operation performed on a group of VMs defined in a runbook. Each operation that a job can run is defined in the migration lifecycle.

Starting a job on a wave

Any job can be started on a wave with a valid runbook. To start a job from the Migrate for Compute Engine Runbook Automation page:

  1. Select your wave.
  2. Click Action > New Job.
  3. Select the required operation as described in migration lifecycle, and then click Start.

Canceling a job

In the event of a migration error, you can cancel a job, which also ends the migration wave. To cancel a job that is in progress:

  1. Select your wave.
  2. Select a row, and then click the Abort button above the table.

    After clicking Abort, the Re-run button appears. Clicking Re-run continues the wave.

Canceling a job stops Migrate for Compute Engine from continuing whatever process was running, but it doesn't revert the Velostrata Manager tasks that are currently running or completed. After resolving any issues, you can resume the job by clicking Re-run.

For example, suppose a wave creates five test clones in five run groups: A, B, C, D, and E. When you click Abort, test clones A and B are already running. Those two test clones continue to run, but test clones C, D, and E are canceled. If you click Re-run, A and B continue to run; the operation doesn't create two new test clones. The wave begins again with a test clone of C followed by D and then E.

What's next?