On October 23, 2026, between 18:39 and 19:40 UTC, customers using our US East region experienced failed requests and elevated latency. A memory accounting error on some of our servers caused them to run out of memory during routine provisioning, which disrupted request handling in that region. The issue has been resolved and we have no indication of data loss. We are making changes to prevent this from happening again.
Resolved
On October 23, 2026, between 18:39 and 19:40 UTC, customers using our US East region experienced failed requests and elevated latency. A memory accounting error on some of our servers caused them to run out of memory during routine provisioning, which disrupted request handling in that region. The issue has been resolved and we have no indication of data loss. We are making changes to prevent this from happening again.
Resolved
Performance levels are back to nominal.
Monitoring
Mitigation in progress; monitoring recovery.
Identified
The team has located the source of the performance problem and has taken measures addressing it.
Investigating
The team is investigating performance alerts impacting US East
