Elevated error rates in internal networking

We've repaired the cluster and are now monitoring the issue.

Emergency maintenance is taking longer than expected. We're still working to address the issue.

We are performing emergency maintenance on the cluster to address the issue. Services may become briefly unavailable.

Cluster issues are preventing our workers from autoscaling. Build queues may be slower than usual.

We've identified a potential issue with our internal cluster networking, and are mitigating the issue while we investigate. This appears to be a continuation of the outage from this morning, which was not fully resolved.

Elevated error rates appear to cause sporadic unavailability to some services. We are continuing to investigate.

We've identified elevated error rates in our internal networking that may impact the stability of some services. We're working to identify the user-facing impact of these errors.