Jobs & queue

The Jobs page in the sidebar shows every active and recently-completed job in Tusk. A job is anything that touches files: a sync, an offload, a restore, a verification pass, a redistribute, a delete-from-everywhere. Each job has a project, a destination (or destinations), a state, progress, and a timestamp.

What you see on the page

• Active jobs: at the top, ordered by most recently started. Live progress, cancel and pause controls.
• Queued jobs: jobs waiting on something (a destination to come back online, the parallelism cap, a dependency on another job).
• Recently finished: a rolling window of completed jobs with their result (success, partial, cancelled, failed) and how long they took.
• Per-project filter: scope the view to one project to focus on what's happening there.

How parallelism works

Tusk runs multiple jobs in parallel without letting them fight each other for the same physical resource. The rules:

• One write per local destination at a time: a single drive can't handle two simultaneous writes well, so Tusk serializes per drive.
• Parallel writes across different destinations: an offload streams to drive A, drive B, and a cloud bucket all at the same time. Each destination is a separate slot.
• Parallel cloud transfers per destination: for cloud destinations, Tusk uses parallel multipart uploads (configurable per-provider) to maximize bandwidth.
• Within-project ingest takes precedence: an active offload pauses the same project's ordinary sync queue so they don't compete.
• Cross-project parallelism: different projects can run their work at the same time without interference, subject to the per-destination caps.

Pause everything

The Pause all button at the top of the Jobs page temporarily halts every active and queued job. Use it when you need full bandwidth for something else (uploading a large delivery, joining a video call on a slow connection). Resume when you're done.

Per-job pause and cancel

Active jobs have inline Pause and Cancel controls. Pause is reversible (resume to continue). Cancel ends the job; partial progress is rolled back where possible (files mid-transfer get cleaned up).

Job result codes

• Success: every file processed without error.
• Partial: some files succeeded, some errored. The job summary shows the error count and offers a Retry action that re-attempts only the failed ones.
• Cancelled: ended by you (or by an internal pause that didn't auto-resume).
• Failed: the job couldn't make meaningful progress. Usually because the destination is offline or credentials expired. The notification panel surfaces the underlying issue with a deep link.

What happens when Tusk quits during an active job

Job state is persisted to disk before Tusk exits. On the next launch, Tusk picks up where it left off. Long-running offloads, restores, and large initial syncs all survive quits and reboots without losing progress.