This incident has been resolved. Between 19:40 and 21:00 UTC, some proxy connections were slow or failed, and the dashboard could load slowly. The cause was an unusually heavy load on our proxy assignment system. We have fixed it, and all services have been working normally since 21:00 UTC. We are sorry for the disruption
No components marked as affected
Resolved
This incident has been resolved. Between 19:40 and 21:00 UTC, some proxy connections were slow or failed, and the dashboard could load slowly. The cause was an unusually heavy load on our proxy assignment system. We have fixed it, and all services have been working normally since 21:00 UTC. We are sorry for the disruption
Monitoring
All affected services have been working normally since 21:00 UTC. We have completed the remaining recovery steps and restored full service to all customers. We are keeping the system under close monitoring and will mark this incident as resolved once we confirm it stays stable.
Identified
Since 19:40 UTC, some proxy connections have been slower than usual or have failed. We identified the cause in our proxy assignment system and applied a fix at 20:45 UTC. The system is starting to recover. Proxy connections may stay slower than usual for a short time. We will post another update within 30 minutes.
Monitoring
The dashboard is loading normally again. Proxy connections are recovering, and some requests may still be slower than usual for a short time. The degraded performance started at 19:40 UTC. We are monitoring the system and will post an update when it is fully resolved.
Investigating
Since 19:40 UTC, the dashboard may load slowly or not load, and some proxy connections are slower than usual or fail. Our proxy assignment system is under heavy load. We have identified the area that causes it, and we are working on a fix. We will post an update in 30 minutes or sooner.