Zum Hauptinhalt springen
This page is not translated yet. Deutsch is switched on so the navigation and URLs can be tested; the text below is the English original. Use the language selector in the header to switch back, or see the documentation home.

Archive and delete jobs

how-to · Reviewed 2026-07-30 GA

Job data accrues storage charges for as long as it exists. Archiving removes a job from your list; emptying the trash is what stops the charges. This page covers the difference.

Archiving does not stop storage charges

Archiving moves output files to the trash. Files in the trash still cost money.

Storage charges continue until files are deleted from the trash. If you archived a year of jobs and are still seeing storage costs, this page explains why.

Archive a job​

From the Jobs list, select the jobs and click Archive.

  • Output files go to the trash automatically.
  • You can also send input files to the trash — worth doing, since input decks are often the larger half.

Archived jobs move out of your default views into the Archived filter. The job record itself is kept.

To archive many jobs together, archive the Workspace Folder holding them: every job and workstation inside is archived in one action, with the same file-handling choices offered here. Note that the control for this is the folder's trash icon, which archives rather than deletes when the folder is not empty.

Archiving also deletes Cloud File System files — permanently

If the job or workstation had CFS mounted, archiving permanently deletes its files from CFS. That side has no trash and no recovery window, unlike Rescale Files, where trashed files survive until purged.

This is on by default, and the files stay gone even if you later unarchive the job. The CLI does the same thing with no warning at all.

Download anything you need from CFS first. See delete CFS data.

Recover an archived job​

  1. Open the Archive folder in the left navigation.
  2. Select the jobs.
  3. Click Recover.
You only get files back if they're still in the trash

Recovering restores the job record either way. But input and output files come back only if they haven't been purged from the trash yet.

Purge happens automatically on a schedule your admin sets, by default, one day after reaching the trash. After that, recovering gives you the job's metadata and nothing else.

One day is a short window. If you might want a job's files back, recover it promptly or download them first.

Delete files directly​

From the Files tab, select files and click the Trash icon. You'll be asked to confirm.

From the Trash folder you can then either:

  • Recover them, while they're still there, or
  • Permanently delete them to stop charges immediately.

Permanent deletion is the only thing that stops storage cost. If you know you're done with a large result set, purge it rather than waiting for the schedule.

Finding what to clean up​

The jobs list has a Storage column showing input plus output volume per job. Sort by it, and export a CSV to total it up.

That's a far better way to find the expensive jobs than archiving by date. A handful of large DOE studies usually dominate.

Admin retention settings​

Workspace admins configure retention under Policies → Platform Settings in the Workspace or Organization Admin portal. Settings apply to all Rescale-managed object storage, whether bring-your-own or the shared account.

SettingDefaultEffect
Auto-move files to trash after a periodOff, 7 days when enabledFiles are trashed automatically
Auto-purge trashed files after a periodOn, 1 dayTrashed files are permanently deleted

With either disabled, files stay recoverable until someone removes them manually.

The default combination is worth understanding

Auto-trash is off, so nothing is trashed unless you archive or delete it. But auto-purge is on at one day, so once something is trashed, you have about a day to change your mind.

The result: nothing disappears on its own, but anything you archive is effectively gone tomorrow.

Before you archive​

  • Download anything you need — see download your results
  • Note the job ID if a support case is open — see share a job
  • Check for shared users who might still be relying on it