What you need to know
- GitHub is mostly back online, but some services are still experiencing problems.
- Git Operations have recovered, while Issues continue to show degraded performance.
- Copilot remains heavily affected, making AI-assisted coding less reliable.
GitHub had a rough Monday. Developers were unable to use one of the world’s most important software platforms for more than three hours, as a widespread outage hit the site. The outage impacted GitHub’s website, code review and merging tools, automated testing and deployment systems, and Copilot.
The problems began around 9:40 a.m. ET and spread quickly. By roughly 11:10 a.m., most of the major GitHub services were down or running poorly. Reports on Downdetector were also flooding in, indicating the issue wasn’t limited to a few users. Reports hit almost 3,000 before slowly falling after GitHub rolled out its fix.
However, there was a fun twist in the GitHub status updates, as error rates for web and API traffic were hovering around 20%, while raw repository content and archive downloads were failing at about 50%. In other words, even getting the code itself might be considerably harder than simply getting to GitHub.
The company said it had identified the problematic component and taken corrective action, with services showing strong signs of recovery. It said at 2:23 p.m. ET that it had mitigated the Git Operations degradation and was monitoring the platform for stability. Copilot is still being investigated and engineers are still monitoring the platform.
AI-powered coding tools are generating exponentially more traffic and workloads, and GitHub has been scrambling to grow its infrastructure. The company had originally planned a 10x increase in capacity, its CTO had said previously, but then realized they needed infrastructure that could handle 30x their former scale.
GitHub has had several incidents this month already, including issues with Actions, Copilot, API requests, login, and Pull Requests. Previous outages have been linked to capacity issues and infrastructure changes, implying reliability is becoming an increasing challenge as GitHub handles ever-more-demanding workloads.
GitHub has not yet publicly stated the exact cause of Monday’s failure. Most services have recovered according to the status page, but the incident remains open.
