Background Processes

Settings, Process Gateway gives you a live view of the helper processes the SHA runs in the background: disk-image export and mount helpers used during restores and file-level recovery, synthetic full backup consolidation, and backup boot-validation sessions. The page has three views - active background jobs, the resources those jobs currently hold, and a recent history of jobs with any errors - plus summary tiles counting active jobs, tracked resources, history entries, and errors.

The page does not auto-refresh. It loads once when you open it, and the Refresh button reloads all three views together, so counts and states are only as fresh as your last refresh.

The Background Jobs table lists the jobs the process gateway currently considers active. Each row shows the job identifier, a status badge, the running process, the supervision scope it runs in, and when it started. When there is nothing active, the table simply reports that no background jobs are running.

Each active job has a Stop action. Stopping asks the job to terminate gracefully and escalates to a forced termination if it does not exit promptly. Once the job stops, the resources it held are released and cleaned up. Stopping a job that has already finished is harmless - the request succeeds without doing anything.

The Reserved Resources table is a read-only list of everything active jobs currently hold on the appliance - for example network ports, attached disk devices, mounted images, file locks, temporary working directories, and recovery containers. Each row shows the resource type and value, the job that owns it, when it was acquired, and the order in which it will be released during cleanup. There are no actions on this table; resources are released automatically when their job finishes or is stopped.

The Job History view shows the most recent 200 jobs, including finished ones: the job type, the Sendense service that submitted it, its status, a short summary of the helper command that ran, how many resources it used, and any error message. The Errors tile counts the history entries that carry an error, which makes this view a quick place to check whether recent restores, consolidations, or validation runs hit problems.

The Restart Process Gateway button restarts the supervision service itself, governed by the Restart when idle toggle, which is on by default.

With the toggle on, a restart requested while jobs are active is accepted and queued: the page shows a Restart queued badge, and the service restarts on its own as soon as no jobs remain active. With the toggle off, a restart requested while jobs are active is rejected with a busy notice telling you how many jobs are still running. Either way, if nothing is active the restart happens immediately.

An idle restart is non-disruptive: when the service comes back, it re-adopts any helper processes that are still alive and cleans up any that are not. Note that the Restart when idle toggle is not saved - it resets to on every time you open the page.

Force Restart restarts the supervision service immediately, even while jobs are active, bypassing both the busy check and the idle queue. The running helper processes themselves are not killed - they keep working and are re-adopted when the service returns. However, starting and monitoring background jobs is briefly unavailable during the restart, and any job whose process happens to die in that window is marked orphaned and cleaned up.

There is no confirmation prompt before a force restart. Prefer the queued restart-when-idle path; keep Force Restart for situations where a normal restart cannot proceed.

Whether the process gateway is available depends on your appliance deployment configuration, not on your license or on any other settings page. If it is not enabled on your appliance, the page shows a notice that the process gateway is disabled instead of the tables.

The page appears under Settings for any role with settings visibility, which includes all built-in roles. For the broader picture of how Sendense runs guarded background workflows, health states, and maintenance windows, see Maintenance And Health; for the performance limits that influence when heavy background jobs are deferred, see Performance Settings.