GitHub rebuilds Git storage architecture from scratch to handle AI traffic surge
Tech

GitHub rebuilds Git storage architecture from scratch to handle AI traffic surge

TechNews Editorial
TechNews EditorialOct 9, 2026 · 2 min read
Share

Why it matters

The new storage architecture aims to restore service reliability and handle massive traffic surges from AI agents without disrupting developer workflows.

The facts

  • GitHub is rebuilding its Git storage architecture from scratch to handle traffic spikes from AI agents.
  • Early internal tests of the new architecture show a 35x write improvement.
  • The new design writes commits to Azure Blob Storage and separates read requests to eliminate bottlenecks.

GitHub is rebuilding its Git storage architecture from scratch to handle unprecedented demand from AI agents. Early tests show a promising 35x write improvement in internal tests.

The plan involves decoupling writes from reads, offloading maintenance, and letting object storage manage redundancy. The redesign happens under the hood with the goal of a non-disruptive change that will not alter developer workflows, review processes, or security controls.

Brian Celenza, a GitHub principal software engineer, wrote in a blog post that enterprise code repositories must support large engineering teams running busy continuous integration pipelines alongside growing fleets of agents. Celenza did not offer a timeline for this migration.

Read nextGitHub updates secret detection model and previews push protection checks

AI agent activity surged on the platform

Platform activity surged as customers shifted their focus to AI agents. Traffic doubled between September 2025 and August 2026, rising from 218.2 billion events per month to 473.3 billion. Commits hit 7.38 billion in September alone, representing a 5x increase from September 2025.

This traffic shift is stressing the system. The GitHub Availability Report states that the service suffered 10 performance-degrading incidents in April. Another nine incidents occurred in May, generating growing public criticism.

GitHub currently uses a storage architecture nicknamed Spokes. Spokes relies on a three-phase commit protocol, storing full repository copies across multiple local disks and requiring a quorum of replicas to acknowledge each write. While this approach ensures strong reliability, it bottlenecks write speed because every push is bounded by the slowest replica required for the quorum.

Several local disks finish recording matching repository copies while one slower disk keeps the shared write operation waiting.
Illustration: AI & Tech News

The new architecture writes to Azure Blob Storage

The new architecture writes a commit only once to Azure Blob Storage. This object store service automatically handles replication and redundancy on its own, eliminating the quorum bottleneck. GitHub also boosts performance by separating read requests into their own channel handled by lightweight compute workers. The only coordination between reads and writes occurs over the reference branch pointers.

Maintenance duties such as compaction and garbage collection now operate as background processes outside the serving path, further reducing system lag. Other companies are also rethinking Git for the agentic age. Former GitHub CEO Thomas Dohmke launched Entire, a Git-hosting service that offloads agent traffic to mirror repositories. SpaceX subsidiary Cursor refined its storage layer by rejecting Spokes' three-phase commit in favor of uploading pushes into an object storage write-ahead log that captures changes as immutable objects while caching at least one copy on fast solid-state disks.

PlanetScale CEO Sam Lambert noted on X that GitHub faces traffic increases far beyond what companies usually prepare for. Brian Celenza did not provide a timeline for the migration.

Newsletter

Get the best AI & tech news daily

A concise daily digest. Unsubscribe anytime.

We use your email only to send this newsletter.

Keep reading