Updated continuously from the same probes that page the on-call engineer. Historical uptime is a rolling 90-day window.
| Component | Region | State | 90-day |
|---|---|---|---|
| Ingest (SRT / RTMPS / WHIP) | eu-waw | Operational | 99.99% |
| Ingest (SRT / RTMPS / WHIP) | eu-fra | Operational | 99.98% |
| Transcode workers | eu-waw | Operational | 99.97% |
| Shielded origin | eu-* | Operational | 99.95% |
| Media stream (gRPC) | eu-* | Operational | 99.96% |
| Control plane API | global | Operational | 99.99% |
Everything customer-visible ends up here, including the ones that were our fault.
12 Jun 2026 · resolved in 41 min. A transit provider re-routed traffic during maintenance and added ~90 ms to origin fetches. We shifted the affected channels to eu-waw and moved the pair back once the path recovered.
28 Apr 2026 · resolved in 2 h 06 min. A queue migration left the retry lane starved, so long-tail encodes queued behind live work. No output was lost; affected jobs completed late and were not billed.