GitHub CTO Vows to Scale Up Before Developers Give Up

On August 17, GitHub experienced a 7‑hour‑47‑minute outage that disrupted operations, pull requests, issues, Copilot, and API services worldwide. CTO Vladimir Fedorov detailed the incident without apologizing, emphasizing that the failure was not caused by recent code or configuration changes but by latent scalability weaknesses.

The outage followed a previous incident on August 6 affecting the Actions platform, indicating ongoing instability. GitHub had already acknowledged in April that its platform could not keep up with growing traffic, despite handling 1.4 billion commits that month and processing 2.9 billion commits, 24 million new repositories, and 130 million pull‑request merges each month.

Microsoft Azure currently carries about 58% of GitHub’s load and half of all Git operations. Fedorov said GitHub is accelerating the migration of more workloads to Azure and plans to build a new architecture whose read capacity scales linearly with the number of readers, starting with the largest monolithic repository.

Before the new architecture arrives, GitHub must address the weak points exposed by retry storms and configuration‑error limits. The company is working to isolate critical systems, tighten retry limits, and add early‑warning signals for traffic spikes.

Social media reactions were mixed: some users sympathized with the challenges of operating at such scale, while paid users expressed disappointment. Fedorov concluded that developers rely on GitHub for building, releasing, and operating their projects, and that trust can only be regained by delivering a more scalable and reliable platform.