Why GitHub’s Massive Outage Happened: Database Infrastructure Rollback Explained

GitHub, the widely used code repository and developer platform, experienced a major outage when its main site displayed an error message indicating that no servers were available, briefly recovering afterward while an angry unicorn image was shown.

During the incident, core services such as Pull Requests, GitHub Pages, Copilot, and the GitHub API were severely impacted, and the GitHub Status page posted its first update at 7:11 PM ET, followed by multiple service alerts.

More than 10,000 users reported the issue on Downdetector, and NetBlocks confirmed an international outage at 7:13 PM ET. Copilot also went down, prompting some users on Hacker News to joke about developers finally being able to "slack off".

GitHub did not immediately comment on the problem. Later, the status page announced that services had returned to normal after a major interruption.

Subsequent updates indicated that the outage was linked to recent changes in the database infrastructure, prompting a rollback of those changes. The article notes that while stateless services can recover easily, stateful databases pose significant challenges when they fail, and the exact database component involved was not identified.

By 8:26 PM ET, GitHub confirmed that the problematic database infrastructure changes had been reverted and that all services were fully operational.

The piece also mentions that since Microsoft’s $7.5 billion acquisition of GitHub in 2018, some users have perceived a decline in service stability.