See related ticket.
We need a way to gracefully abort some jobs.
It runs commands to wipe a disk with "dd"
If we abort the job with core.job_abort the job will abort but dd will keep running.
We need a way to make that clean and actually abort/kill the dd command too.
We will likely need a flag in the job metadata to let the API user know a job can be safely aborted (so the UI can know to which jobs the core.job_abort can be used).