GitHub October 2025: Four Incidents, One Lesson

TL;DR

• GitHub had four incidents in October affecting Actions, Codespaces, API, and third-party integrations
• Longest outage lasted nearly 7 hours; most were caused by configuration errors or external dependencies
• Check the status page for details; GitHub is reducing reliance on external providers

What Dropped

GitHub published its October 2025 availability report, documenting four separate incidents that degraded services across the platform. The incidents ranged from 1 hour to nearly 7 hours, affecting everything from Actions runners to Codespaces to the Enterprise Importer.

The Dev Angle

October 9: A network device brought back into production before repairs finished caused packet loss. API users saw up to 7.3% error rates initially. Actions runs were delayed by an average of 13 minutes, with 24% of runs affected. GitHub is tightening validation on device repairs to prevent this.

October 17: Mobile push notifications failed for 70 minutes due to a bad configuration change to cloud resources. This hit both github.com and GitHub Enterprise Cloud globally. The team is reviewing cloud resource management procedures.

October 20: Codespaces degraded for 2+ hours when a third-party dependency for building devcontainer images went down. Creation errors hit 71% at peak; resume operations averaged 23.4% failures. GitHub is working to remove this dependency from the critical path.

October 29: The worst incident—a third-party provider outage cascaded across multiple services for nearly 7 hours. Codespaces hit 90% error rates (100% at peak). GitHub Actions large runners saw 10% of jobs fail or delay beyond 5 minutes. Enterprise Importer migrations failed. Copilot Metrics API downloads were unavailable (~100 failed requests). GitHub is now prioritizing reducing critical dependencies on external providers and implementing graceful degradation.

Should You Care?

If you rely on Codespaces for development, Actions for CI/CD, or GitHub Enterprise services, October was rough. The October 29 incident was particularly painful—a 7-hour outage with 90%+ error rates on Codespaces is not a minor blip.

The pattern is clear: GitHub's biggest vulnerabilities are external dependencies and configuration management. If you're building on top of GitHub services (like using Copilot for code review or relying on Actions for deployment), you should have fallback strategies for extended outages.

The good news: GitHub is being transparent about root causes and committing to architectural changes—removing critical external dependencies and improving monitoring. That's the right move. Watch their status page for real-time updates and post-incident recaps.

Source: GitHub Blog