GitHub Actions and Pages Experience Degraded Availability, Migration to Azure Accelerated

Outage Affects GitHub Services

On August 6, GitHub experienced an incident that led to degraded availability for multiple services, including GitHub Actions and GitHub Pages. This resulted in workflow runs failing or being delayed, and some queued jobs potentially timing out.

The incident also impacted other GitHub functionalities, such as self-hosted runners, which may have seen errors or rate limiting during registration. Services like Copilot code review, Copilot coding agent, hosted runners, and migrations using GitHub Enterprise Importer were also affected. Webhook deliveries experienced delays.

GitHub Acknowledges Impact and Shortcomings

GitHub stated that the incident on August 6 was unacceptable in its impact and duration, acknowledging that it fell short of its availability commitments to users. The company recognized the significant impact such a prolonged outage has on user productivity and trust.

A deeper root cause analysis (RCA) is underway to fully understand all contributing factors. GitHub plans to provide a complete update in its August availability post, to be published in September.

Accelerated Migration to Azure

In response to the incident, GitHub is accelerating its architectural roadmap for GitHub Actions. This roadmap focuses on improving isolation, resiliency, and scalability for the service.

A key part of this acceleration involves migrating the GitHub Actions service fully to Azure. GitHub noted that the core Actions service was still running in its data centers during the incident, which was a contributing factor to capacity issues.