Broken provider? No problem.
If a provider returns 5xx errors or timeouts, it gets isolated automatically — the other 28 keep running. After a short pause, Sublarr quietly retries it.
Quotas protected
Daily limits are tracked. Multiple API keys per provider? Sublarr rotates them before any single one hits the wall. No "429 Too Many Requests" at 3 AM.
Restart-safe
Container update? Power loss? Doesn't matter. Pending searches, scheduled jobs, and provider state survive restarts — exactly where they left off.
You stay in the loop
In the Scheduler tab you see every background job: when it last ran, what's queued next, whether anything failed. Pause, run on demand, or change the interval — all in one click. Prometheus metrics live at `/api/v1/metrics` if you want to wire them into Grafana.
Trust the scanner — or watch it
Both work. Sublarr is just as solid as a set-and-forget tool as it is part of a stack with a Grafana dashboard.