API connectivity degradation for a subset of users in the Ireland region.

Summary Between 2026-01-22T09:33:35.957Z and 2026-01-22T15:08:57.765Z today, a subset of API requests originating from the EU West (Dublin) region failed to connect to our backend services, resulting in timeouts or connection errors (5xx). Traffic from other global regions (e.g., France, US, Asia) remained unaffected.

The issue was traced to a regional routing conflict between our edge network provider and our application hosting infrastructure.

A background infrastructure update in the Dublin datacenter created an incompatibility in how internal traffic was routed between these two providers within the same physical location. This caused connections initiated from Dublin to be dropped before reaching our application servers, despite the services being healthy.

Resolution

Our engineering team identified the routing loop and deployed a configuration update to force traffic through an alternative, stable network path. This effectively bypasses the conflict in the Dublin datacenter.

The fix has now been verified, and all systems are operating normally. We have implemented monitoring to ensure this specific routing path remains stable.