GitHub experienced a global outage of its various services due to a database infrastructure-related change that was rolled back. The Microsoft-owned code-hosting platform started experiencing issues just after 2300 UTC (1600 PT) and affected multiple public services, including GitHub Actions, pages, issues, pull requests, Copilot, Codespaces, packages, Git operations, and webhooks.
The website displayed a unicorn and error message, stating that no server was available to service the request. Users were unable to access GitHub.com, the GitHub API, or SSH-based access to repositories.
In an update, GitHub announced it had decided to roll back the infrastructure change and was monitoring for full recovery. By 2345 UTC, the site started to return to normal, with users reporting improvements in service health.
It’s not the first time a major technology company has experienced issues due to changes gone awry. In this case, GitHub’s database infrastructure update caused the outage, but the platform is now recovering and back online.
Source: https://www.theregister.com/2024/08/14/github_rollback/