GitHub: monthly commits doubled in four months, overwhelming server capacity
GitHub says Monday's outage, which knocked out several services for more than seven hours, was caused by a capacity failure rather than a code or configuration error, after monthly commits doubled from 1.4 billion to 2.9 billion since April.
GitHub has explained what caused several of its services to go down for more than seven hours on Monday, saying the outage stemmed from a capacity failure rather than any error in code or configuration. According to the company, traffic hit a new peak that a key infrastructure component in its Central US data center could not scale to match.
A sharp rise in activity
GitHub disclosed that monthly commits have grown from 1.4 billion in April to 2.9 billion now, more than doubling in just four months. The company said this growth explains the pressure placed on its systems, though it does not excuse the outages that resulted. The surge lines up with the growing volume of code being generated with the help of artificial intelligence.
Fixes being rolled out
On its status page, GitHub provided a more technical account of what went wrong. The main fixes it is implementing focus on how the platform handles simultaneous activity across its services. Specifically, GitHub is reviewing CPU and memory alerts to spot server components that might fail under heavy traffic, and it is introducing consistent retry limits, retry budgets, and variable timeouts for interactions between services, aiming to prevent so-called retry storms and cascading load failures.
These measures address GitHub's immediate technical problems but do not resolve the underlying trend driving the traffic spike: an increasing share of code being produced by AI tools. With Anthropic and OpenAI both continuing to expand their AI coding products, the pressure on GitHub's infrastructure is likely to persist, given its role as the primary home for both professional and hobbyist software projects.
