Ask HN: GitHub employees what's going on? Why?

The Hacker News Ask HN post shows developers seeking direct explanations from GitHub employees about recent platform issues. Multiple commenters point to AI-assisted coding as the driver behind a sharp rise in commit activity. They cite a 14× increase in commits over the past year and note that GitHub recorded 1 billion commits in 2025. The community wants to know whether this growth is outpacing the service’s capacity.

For engineers who depend on GitHub for source control and CI/CD, any degradation translates into extra work and possible delays. If the service cannot handle the current load, teams may need to evaluate self-hosted Git solutions or other forge platforms. Such a shift involves migration effort, retraining, and potential loss of integrated GitHub Actions, Packages, and security alerts. Thus the cost of adopting an alternative includes both engineering time and possible reduction in workflow automation.

Commenters warn that the platform may start to exhibit slowdowns, timeout errors, or failed build jobs when commit traffic exceeds its handling threshold. They note that Azure, the underlying cloud provider, has been described as unreliable in past discussions. When the infrastructure struggles, pull request merges can be delayed and automated tests may not run on schedule. This directly impacts release cadence and increases the overhead of troubleshooting intermittent failures.

Opinions differ on the root cause, with some attributing the problem to Azure’s performance limits and others to GitHub’s software architecture. One commenter argues that Microsoft’s investment in data centers matches that of OpenAI, implying the issue is not pure hardware capacity. Another contends that the software stack itself is unstable and needs architectural changes to cope with the growth. The lack of consensus means engineers must monitor the situation and prepare contingency plans rather than rely on a single guaranteed fix.