GitHub availability report: September 2026

GitHub’s August and September availability work focused on reducing shared-database load, adding overload protections, improving GitHub Actions capacity, expanding pull request failure containment, and migrating production workloads to Azure, though uneven Azure availability-zone capacity paused deployments and caused the September 30 service-migration target to be missed. By October 7, 120 of 160 production services and 13 of 51 primary MySQL databases had moved to Central US, which was serving substantial portions of core application, Git read, and stream-processing traffic. Database improvements removed about one million queries per second from an older shared database, deleted 980 million unnecessary edit-history rows, reduced weekly deadlocks by about 77%, and enabled limited pull request functionality during some shared-dependency failures. GitHub also reduced Actions Redis operations and CPU usage while adding caching that sharply lowered bandwidth, latency, and queued requests. The report details a September 13 incident in which a background cleanup job exhausted connections on a shared primary database, affecting issues, pull requests, APIs, token creation, Actions, and signups; GitHub paused the job and plans stronger capacity-aware safeguards, timeouts, retry controls, and recovery behavior. A separate September 28 incident delayed Copilot code reviews after a rollout sent incomplete results for some users, and GitHub reverted the change while planning improved delivery safeguards, automated recovery, monitoring, and deployment processes.

Use this post, company, and trend context to find content marketing opportunities, perform competitive analysis, or address product feature gaps via the Plushcap MCP server or the Plushcap API.