GitHub's March 2026 availability report details four incidents causing degraded performance. The first (March 3) was triggered by a bug during a caching mechanism deployment that caused mass cache expiration, leading to ~40% request failures on github.com and ~21% errors on Copilot. The second (March 5) saw GitHub Actions severely impacted for nearly 3 hours due to incorrect Redis load balancer configuration during infrastructure updates. The third (March 19-20) involved the Copilot Coding Agent going down twice due to a system authentication issue preventing datastore connectivity, with error rates peaking at 100%. The fourth (March 24) affected Microsoft Teams integrations for nearly 3 hours due to an upstream dependency outage. GitHub outlines remediation steps for each incident including killswitches, improved monitoring, automated credential lifecycle management, and configuration change safeguards.
Sort: