Skip to content

Maintenance

Tab: System → Maintenance.

Requires: System admin (Maintenance permission).

Operational dashboards for host resources, storage, database health, active sessions, log downloads, and service restarts.

System Maintenance tab

Restart services

Control Effect
Restart engine Restarts the GPU/engine process. Confirm in the dialog; wait for the overlay until the command and events channels are ready again (timeout on the order of several minutes).
Restart Web service Restarts the API/web process. Confirm; wait until the API responds again (shorter timeout). The UI reconnects realtime after the bounce.

Tip

After uploading a custom module or changing settings that require an engine reload, use Restart engine.

Engine restart may be disabled while the command channel is already down. During a planned bounce, temporary ERROR / STOPPED states are expected until the service is healthy again.

Host resources

Element Description
Host resources Card title
Engine status badge Current engine state (for example RUNNING, STOPPED)
Application data disk Disk usage for application data
Engine detail Active/queued pipelines, max pipelines, communication channels
Health Engine health metrics when available

Data refreshes on an interval while the tab is open. If a refresh fails, the card shows the last successful snapshot with a warning.

Storage by source

Summary tiles for total sources, system DB media size, and upload staging usage, plus per-source breakdown when data is available.

Database status

MongoDB health: uptime, connections, data/index/total sizes, and collection statistics.

Active sessions

Lists users currently connected to the realtime channel. Shows an empty state when nobody is connected.

Logs

Download rotated service logs as a zip archive.

Field Description
Service All services, API only, or Engine only
Hours to include 1–168 hours
Download logs Starts archive download

Possible errors: no files in the time window, archive too large (shorten the window), or rate limiting.