Resolved

2025-10-06 09:23 UTC
We identified the root cause to be intermittent network disruptions on a NAT instance, which resulted in blocking multiple workers from accepting new requests due to excessively configured timeout values. We have implemented improvements to both database connection configuration and worker timeouts and have put additional monitoring in place.

Investigating

2025-10-05 10:08 UTC
We continue investigating this past issue and have put additional monitoring into place.

Investigating

2025-10-05 07:01 UTC
We are investigating ping endpoint failures that occurred between 00:30 and 01:20 UTC.