GitHub Pull Requests Hit 7.5 Hour Outage with 500 Errors March 31

What happened

GitHub's Pull Requests service experienced a 7.5-hour incident on March 31, 2026, from 13:53 UTC to 21:23 UTC. The service suffered elevated latency and failures with an average error rate of 0.15%, peaking at 0.28% of all requests returning 500 errors. GitHub attributed the incident to a change in garbage collection settings on their servers. The issue was resolved by 21:23 UTC the same day.

Business impact

Enterprise teams relying on GitHub for code review and deployment workflows faced potential delays in UAT cycles and production releases during the 7.5-hour window. Organizations with same-day deployment schedules or critical hotfixes could have experienced bottlenecks in their testing and approval processes.

Background

GitHub serves as the primary code repository and collaboration platform for most enterprise development teams, making Pull Request functionality essential for code review workflows. Even sub-1% error rates can significantly impact enterprise teams during peak development hours when multiple UAT cycles and deployment approvals are running simultaneously.

What this means for your team

Review your deployment and UAT schedules to identify critical dependency windows on GitHub services. Implement backup code review processes or local Git workflows that can function during GitHub outages. Monitor GitHub's status page during planned releases and consider delaying non-critical deployments when incidents are reported.

What to watch

Monitor GitHub's status page for any follow-up incidents related to garbage collection configuration changes. Track whether similar performance issues emerge in other GitHub services that could affect CI/CD pipelines.

Sources

Incident with Pull Requests: High percentage of 500s
GitHub Status