Why GitHub’s Recent Outages Happened: Capacity Limits and Scaling Challenges

Outage cause

GitHub experienced a prolonged large‑scale outage. The post‑mortem identified a sudden spike in usage that triggered internal infrastructure errors; existing capacity constraints amplified the impact. No code or configuration changes were involved.

Traffic growth data

Monthly commit volume was roughly 5 × 10⁸ in 2023‑2024, rose to 6‑7 × 10⁸ in early 2025, and surged to 2.9 × 10⁹ by August 2026. Pull‑request and new‑repository creation rates also reached new highs during the same period.

Capacity expansion response

Added approximately 3 million CPU cores and 120 PB (≈120 000 TB) of high‑speed storage in the current year.

Accelerated migration to Microsoft Azure; Azure now handles about 58 % of GitHub’s platform load and 50 % of Git operations, up from roughly 12 % in May.

Increased staffing for availability, assigning more teams to testing, secure deployment, observability, and alerting mechanisms.

Current status

Progress on availability has been made, but work remains and further time and effort are required.

Code example

往
期
推
荐
1、
被Linux之父力挺的软件，开源后倒下了...
2、
公司招了个五年开发经验的程序员，但是感觉是个萌新，怎么办？
3、
用 AI 写代码更快了，为什么反而更累了～
4、
“宁愿失业，也不再用AI写一行代码！”用了18个月AI后，20年老兵宣布停用：要重获手搓代码的快乐
5、
当公司开始实施“推恩令”，你就该离职了
点
分
享
点
收
藏
点
点
赞
点在看