GitHub Outage History

Users worldwide experienced delays in GitHub Projects, with changes to projects not reflecting immediately. The disruption began at Apr 21, 2026 3:06 PM UTC and led to a backlog of updates while GitHub investigated and implemented mitigations, causing delays in project changes and related workflow updates.

GitHub communicated an ongoing investigation and deployed fixes to relieve the processing backlog. As the team worked to clear pending events and restore normal timing, recovery progressed and the status was updated to up at Apr 22, 2026 1:26 AM UTC as the backlog cleared and updates resumed typical timing; some residual delays may have persisted during the transition.

Developers relying on GitHub's code scanning and code quality checks for pull requests faced degraded performance beginning 1:31 PM UTC as GitHub flagged a partial degradation in code scanning defaults and code quality features. PR scans and quality analyses were not reliably triggered on new pull requests, impacting typical review workflows. User reports circulated through the afternoon, signaling a broad impact on PR-related tooling and project-management signals.

Mitigations progressed over the next several hours. The incident escalated to a more disruptive state at 2:01 PM UTC, with the root cause identified and mitigations being deployed. By 4:21 PM UTC, code scanning and Code Quality triggers were back up and running, but developers were warned that PRs created before or during the incident might require a new push to trigger analyses, and that some new issues were not showing on project boards. Deployments and remediation continued through the evening, with the situation described as mitigated by 6:21 PM UTC and ongoing monitoring. A note at 3:11 AM UTC indicated that issues linked to projects during the incident could take roughly three more hours to render while a re-index completed, and stability was rechecked by 4:21 AM UTC before final recovery. The service returned to up status by 5:06 AM UTC.

GitHub.com users faced slow page loads and failed requests as roughly 10% of web traffic was affected. The disruption began around Apr 17, 2026 2:18 PM UTC and manifested as slow responses and errors across core web endpoints, with reports of timeouts when loading pages or performing actions like repository access. A surge in user reports mid-afternoon UTC reflected the growing impact. StatusGator detected the issue about 43 minutes before GitHub publicly acknowledged it. The root of the disruption was traced to a problematic component, which GitHub proceeded to isolate as part of mitigation.

Users of GitHub Codespaces faced degraded performance when creating new Codespaces or starting existing ones from the VS Code editor, with SSH connections remaining unaffected. The disruption began around 3:11 PM UTC and rapidly escalated to a broader outage, affecting about 70% of Codespaces by 3:51 PM UTC. Reports described sessions failing to start and, in some cases, complete outages for parts of Codespaces users, including observations from Europe and North America.

GitHub Copilot users across the globe faced language model unavailability, causing Copilot to fail to generate code suggestions and other Copilot features to become unresponsive in IDEs and chat. The language model powering Copilot was unavailable for much of the day, with reports of 'Language model unavailable' and related errors affecting users in multiple regions. Impact began around 5:15 AM UTC and persisted for most of the day, with recovery by 9:47 PM UTC.

Users relying on GitHub Copilot experienced disruption as the Copilot Dashboard on the Insights tab became inaccessible. The issue appeared to affect Copilot-related services rather than the entire GitHub platform, with reports of features not loading and language-model services unavailable across multiple regions. Impact began amid early user reports around 1:45 AM UTC and expanded as updates acknowledged the Copilot dashboard issues.

GitHub’s status updates escalated the incident, first noting disruption with some GitHub services at Apr 14, 2026 2:01 AM UTC. By Apr 14, 2026 4:41 AM UTC the Copilot Dashboard on the Insights tab was identified as the affected component and mitigation efforts were underway. The issue gradually stabilized, with the status returning to normal around 6:11 AM UTC.

A brief outage affected GitHub Pages, degrading the availability of websites hosted on the Pages service. Users attempting to load Pages-hosted sites experienced failures or delays. The incident began at Apr 13, 2026 7:56 PM UTC and was mitigated by 8:31 PM UTC, with full restoration by 8:36 PM UTC as GitHub worked to ensure stability.

GitHub’s status updates indicated the Pages degradation and subsequent recovery, with officials monitoring the service to confirm stability after mitigation.

GitHub users worldwide faced intermittent performance and access issues impacting Copilot and other GitHub services. Impact began around 4:02 PM UTC per StatusGator's Early Warning Signal. GitHub acknowledged the disruption around Apr 13, 2026 4:46 PM UTC and said the root cause had been identified, with a Copilot fix being rolled out and full recovery expected in 5 to 10 minutes. StatusGator’s early warning ended around Apr 13, 2026 5:08 PM UTC as services began to recover, with subsequent updates indicating progress toward full restoration.

GitHub Copilot experienced a brief degraded performance affecting Copilot Agent Sessions (CCA/CCR). The incident began at Feb 25, 2026 4:41 PM UTC and was resolved by Feb 25, 2026 4:47 PM UTC, lasting roughly six minutes. The GitHub Status page listed Copilot as degraded during the window before returning to normal. The provider issued two status updates describing the issue and its resolution.

The GitHub code search service experienced degraded performance starting at 8:01 PM UTC and was resolved by 12:51 AM UTC, lasting about 4 hours and 50 minutes. Users attempting to search for issues or pull requests could encounter slower results or timeouts, with code search latency impacting related searches. The issue was traced to a single shard with elevated CPU usage, causing degradation across code search and later affecting searches for issues and pull requests.