Multiple services have elevated errors and endpoint failures when checking feature flags
This summary is created by Generative AI and may differ from the actual content.
Overview
Approximately 83% of requests to the analytics endpoint serving the /chronicle feature failed between 17:38 UTC and 18:22 UTC on June 15, 2026, due to an internal feature-flag service that encountered a transient error and failed to recover.
Impact
Approximately 83% of requests to the analytics endpoint failed, resulting in degraded functionality for customers.
Trigger
An internal feature-flag service encountered a transient error and failed to recover.
Detection
The issue was identified through reports of impacted performance for some GitHub services.
Resolution
The service health was restored by removing the feature flag gating the analytics endpoint and deploying that change. Additionally, the feature-flag client was changed to retry errors that are not known to be permanent.
Root Cause
The root cause was a transient error in the internal feature-flag service that caused feature flag checks to fail, resulting in requests being rejected.