This incident has been resolved.

This incident has been resolved.

The scheduled maintenance has been completed.

Postgres services that have had maintenance in the last 8 days (since 2025-06-18) may be missing some, but not all, metrics during this period (e.g. active connections). We have implemented a fix and services will have maintenance scheduled to apply the fix. Metrics will resume after maintenance for a service is complete but metrics that are already missing are not able to be recovered.

This incident has been resolved.

Workloads were delayed in starting for a subset of users in the Oregon region. This includes builds, crons, and service scale-ups.

Between 2025-05-07 and 2025-05-13, services may have experienced an increase in instances being evicted and restarting. This was due to a failure of a routine cleanup task, which failed in a way that did not trigger our monitoring. We have fixed that task as well as improved monitoring and alerting to prevent this from recurring.

This incident has been resolved.

Due to an issue with our infrastructure provider, some services may have experienced downtime between 21:58 and 22:04 on 2025-05-04. Builds and deploys may have been impacted during this time.