Refreshed every minute. Ninety days of incident history.
| Component | State |
|---|---|
| Ingest | operational · 79 ms |
| Packaging | operational · 0.5× realtime |
| Delivery — Europe | operational · 14 ms |
| Delivery — Asia | operational · 31 ms |
| Reporting | degraded · aggregation behind |
Hourly figures are roughly twenty minutes behind. Delivery and playback are unaffected — this is the dashboard and the reporting endpoints only. The backlog is reprocessing and we expect it clear within the day.
A run of long masters filled the transcode queue for about forty minutes. Nothing was lost; everything queued eventually completed. We have since capped per-account concurrency so one large batch cannot crowd out everyone else.
One delivery region returned errors for roughly six minutes before health checks drained it. Players failed over as designed; a small number of sessions saw a stall.