Aller au contenu principal
This page is not translated yet. Français 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.
how-to · Reviewed 2026-11-02 GA

Activity & job administration

The Activity section of an admin portal is the only place you can see and act on other people's work. Everywhere else in the platform a member sees their own jobs; here an administrator sees all of them, and can shut them down, delete them, or reassign them to a project.

Both portals have Activity, and the scope differs: the organization portal covers every workspace, a workspace portal covers one.

Activity pageOrganizationWorkspace
Jobs✅✅
Workstations✅✅
Storage Devices✅✅
Clusters✅—
Cloud Filesystems—✅

The admin job list​

Activity > Jobs lists every job in scope, not just running ones. On a busy workspace that is a large list, so the filters matter more than the table.

Two views​

A selector at the top switches between:

  • Full Jobs List — everything, at any status.
  • Pending Jobs Queue — only jobs waiting to start, in the order they are queued.

The queue view is the one to reach for when members report that work is not starting. For what a job is actually waiting on, see why a job is queued.

Filtering​

Filter by Id, Name, Owner or Project, and by two separate status controls:

  • Status: Any, Saved, Starting, Executing, or Completed.
  • Deleted: Any, or Deleted.

The deleted filter is the useful one that is easy to miss. Deleted jobs are hidden by default, so if you are investigating something a member says they removed, switch it to Deleted or you will conclude the job never existed.

The status filter offers fewer states than the list displays

Those five options are everything the Status filter has, but the Status column shows states that are not among them, including Pending and Deleted. So there are jobs you can see on the page and cannot filter for.

The one that matters in practice is Pending, because a pending job is usually exactly what you are looking for. Use the Pending Jobs Queue view above instead of the status filter, which is what that view is for.

The portal's vocabulary is also its own. It says Saved and Executing where the member-facing platform and the job statuses reference call the same states Draft and Running, so filter on what the portal shows rather than on the status a member reported.

What the columns tell you​

ColumnNotes
Name, IDThe job's name and its Rescale ID. Quote the ID in support conversations, since names are not unique
StatusThe job's current state, including Pending and Completed. A deleted job reads Deleted here, which is how one shows up once you have switched the Deleted filter on
ReasonsWhy the job ended the way it did, for example "A run failed" or "User terminated". This column is the fastest triage signal on the page
OwnerThe member who submitted it
CoresCore count and coretype, for example 8 x Kyanite
WorkspaceWhich workspace the job belongs to, linking through to it. Organization portal only, since a workspace portal already covers exactly one
LineageWhere the job came from, naming the source and its ID: Duplicated from another job, or Created from template for one launched off a template. Blank for a job built from scratch. This is what lets you trace a bad configuration back through every copy of it
CreatedSubmission time
ProjectThe assigned project, blank if none

Bulk actions​

Select jobs with the row checkboxes, then choose from Select Action:

Shut down jobs. Stops running jobs. Use it to reclaim capacity from work that is clearly wrong, or to clear a workspace that is up against its core limit. Compute charges stop when the job stops, so this is also the lever for a job burning money on a bad setup.

Delete jobs. Removes jobs, and it is also how a job template is deleted. Templates sit in the job list alongside ordinary jobs, marked with a TEMPLATE badge against the name, so select the template and choose Delete jobs; a dialog asks you to confirm, and the template is removed once you submit it. Either an organization or a workspace administrator can do this. Template names are often reused, and a job launched from a template inherits its name, so filtering by the template's ID is the surest way to act on the one you actually mean rather than on something created from it.

Members cannot do this at all. They can only archive a template from the template gallery, which hides it without removing it; deleting one outright is an administrator action taken here.

In a single-workspace organization, look in the organization portal

The workspace job list normally sits under Activity in a workspace admin portal. In an organization with only one workspace it appears under Activity in the Organization Admin portal instead, which is where to go to delete a template there.

Assign jobs to a project. Moves jobs into a project. Two common uses: correcting work submitted under the wrong project, and bringing historical jobs under a project so they appear in its billing. The same action exists on the project page as Assign jobs to this project.

These actions apply to other people's work

Shutting down or deleting affects members who have no visibility into why their job disappeared. Deletion in particular is not something they can undo. Tell the owner, or filter by Owner first so you are certain whose work you have selected.

Workstations and storage devices​

Activity > Workstations and Activity > Storage Devices are the equivalent admin lists for interactive sessions and storage. Both are commonly the answer to "what is still running and costing money", since a forgotten workstation bills for as long as it is active.

For what these resources are and how members use them, see workstations and storage devices. For the settings that limit them, see auto-terminate suspended workstations and storage and transfer limits.

Clusters​

Activity > Clusters is organization-only. It lists the underlying compute clusters, filterable by status: Pending, Starting, Started, or Stopped.

It is a low-level view, and most cluster questions are better answered from the job that owns the cluster. See cluster health for diagnosing a specific job's cluster.