Platform: US - Login Authentication Errors

**Incident Overview** 

On May 10, 2026, between 23:47 UTC and 02:48 UTC \(the next day\), some US customers in the East US region experienced intermittent login errors, including 504 timeouts and slow response times. The issue originated at the Ashburn, VA location of our web application firewall \(WAF\) infrastructure, which handles traffic routing for East US region. Service was restored during the incident by rerouting traffic through alternate locations. The Ashburn site was returned to service at 12:00 UTC on May 11, 2026. 

**Root Cause** 

The outage was triggered by a hardware failure within the Ashburn WAF infrastructure. A faulty optical transceiver caused the primary network link between aggregation switches to become unstable, leading to repeated connection drops. A secondary link that should have served as a failover was already inactive at the time, a condition that had gone undetected due to a gap in infrastructure monitoring. With no functioning backup, the degradation of the primary link disrupted traffic flowing through the site, resulting in the login errors customers experienced. 

Two factors extended the duration of impact. First, the monitoring gap meant that the inactive backup link was not visible ahead of the incident. Second, there was a delay in executing the standard failover process \(deactivating the affected site and rerouting traffic\) once customer impact was identified. Both gaps have since been identified and remediated. 

**Preventive Actions** 

* The faulty optical transceivers on both the primary and secondary network links at the Ashburn location are being replaced, with a spare unit kept on-site to support faster response if needed in the future. 

* The monitoring gap that failed to surface the inactive secondary link is being addressed, ensuring both active and backup links are visible to the network operations team going forward. 

* Response procedures are being reviewed and reinforced to ensure the site deactivation and traffic rerouting process is initiated promptly whenever significant customer impact is detected.

Between approximately 7:47 PM and 10:48 PM EDT on Sunday, May 10, 2026, some US customers in East US region may have experienced intermittent timeout errors (including 504s) or delays when logging in to Platform.

This was caused by a disruption at a US datacenter operated by our upstream network provider, which affected traffic routing in the region. The upstream issue has since been resolved and affected infrastructure has been restored.

Platform login is now operating normally, and we are continuing to monitor. We apologize for any inconvenience this may have caused.