Update: We've refined the impact window for the July 23, 2026 incident based on a detailed log review: customer-facing impact began at approximately 08:21 UTC (earlier than first reported) and was resolved at 14:31 UTC. Our RCA has been updated accordingly.

Incident Summary

Between the hours of 08:21–14:31 UTC on July 23, 2026, some customers were unable to log in due to intermittent service failures.

Affected: All customers.

Symptoms: Users reported inability to login.

A recent infrastructure dependency upgrade caused the service to exhaust system resources under load. This led the service to intermittently stop accepting connections, which in turn caused login requests to fail.

We rolled back the change and redeployed the service across all production environments. Service stabilized after the rollback, and we continued monitoring to confirm the issue did not recur.

We are implementing additional safeguards, including tightening dependency version controls, fixing underlying resource-handling issues in the service, and adding monitoring to detect abnormal resource usage before it impacts customers.

The issue has been fully resolved. All systems are operating normally.