GitHub Actions wobbled for nearly four hours

On the evening of 5 October 2026, GitHub suffered a prolonged outage in its Actions service. According to the official status page, the incident began at 19:11 UTC and was only closed at 22:49 UTC — nearly four hours later.

The first reports concerned delays in assigning GitHub-hosted runners. As a result, workflows started late across various runner configurations: jobs either sat in the queue or took noticeably longer than usual to begin. The engineering team said it was investigating the cause and working to mitigate the impact.

By 21:09 UTC it became clear the problem was broader than it first appeared. Beyond Actions and hosted runners, some users lost access to their repository lists, as well as to licensing and billing pages within their accounts. GitHub Pages was then added to the list of affected services, suffering degraded performance of its own.

The turning point came around 21:32 UTC, when GitHub applied mitigations for the Actions failure. The backlog of queued jobs began to clear, new jobs stopped being delayed, and the company expected full runner recovery shortly. Work on the other affected services continued in parallel.

By 21:54 UTC Actions was operating normally, and by 22:40 UTC Pages had recovered as well. The final note at 22:49 UTC confirmed the incident was resolved. GitHub promised a detailed root cause analysis later, once it is available.

Formally the outage affected two components — Actions and Pages — but in practice its reach was wider: from CI/CD pipelines that depend on automated builds to everyday operations around billing and repository management.

Source: Hacker News