Incident with Git Operations
Facts
- Source
- GitHub Status · original
- Started
- (10 days ago)
- Ended
- · lasted 1 h 44 min
- How it ended
- Microsoft marked it resolved.
- First captured by us
- Last changed
- (43 min ago)
- Affected services
- GitHub Git operations
Last text from Microsoft
On November 3, 2026, between 15:00 UTC and 22:37 UTC, Git operations over HTTPS on GitHub.com were degraded. Clones, fetches, and pushes over HTTPS were slow, and some failed. For about half of Git HTTPS traffic, median response time rose from under 100 milliseconds to as high as 5 seconds. On average, the error rate was 0.65% and peaked at 1.29% of requests to the service. Git operations over SSH were not affected. <br /><br />This was due to a buildup of excess connections on the hosts that serve Git HTTPS traffic, which saturated their network capacity and caused slow and failed responses for all clients served by those hosts. <br /><br />We mitigated the incident by blocking a client traffic pattern at our load balancers and reloading them, which released the excess connections. The block also prevented some users of one third-party integration from performing Git operations over HTTPS until we removed it on November 4 at 13:26 UTC. <br /><br />We are working to alert on sustained Git HTTPS latency sooner and make load balancer mitigations take effect on existing connections immediately.
Quoted from GitHub Status (source), captured . Reproduced for reporting and archival purposes.
Official updates
Posted by GitHub Status, newest first.
Resolved
On November 3, 2026, between 15:00 UTC and 22:37 UTC, Git operations over HTTPS on GitHub.com were degraded. Clones, fetches, and pushes over HTTPS were slow, and some failed. For about half of Git HTTPS traffic, median response time rose from under 100 milliseconds to as high as 5 seconds. On average, the error rate was 0.65% and peaked at 1.29% of requests to the service. Git operations over SSH were not affected. <br /><br />This was due to a buildup of excess connections on the hosts that serve Git HTTPS traffic, which saturated their network capacity and caused slow and failed responses for all clients served by those hosts. <br /><br />We mitigated the incident by blocking a client traffic pattern at our load balancers and reloading them, which released the excess connections. The block also prevented some users of one third-party integration from performing Git operations over HTTPS until we removed it on November 4 at 13:26 UTC. <br /><br />We are working to alert on sustained Git HTTPS latency sooner and make load balancer mitigations take effect on existing connections immediately.
Monitoring
The degradation affecting Git Operations has been mitigated. We are monitoring to ensure stability.
Investigating
We are seeing significant improvement, though some Git Operations may be slow for a small set of customers. We are marking this as mitigated and will continue to monitor.
Investigating
Our mitigation is rolled out and customers should start seeing improvements to Git Operations. We will continue to provide updates on mitigation status.
Investigating
We have identified the issue impacting Git Operations and are beginning to roll out a mitigation. We will follow up with another update when we begin to see recovery.
Investigating
We are investigating reports of degraded performance for Git Operations
What Microsoft changed
Microsoft edits the text of an incident in place. We keep every version we capture (checked every minute) and highlight the difference. Removed text is struck through, added text is marked.
First version we captured
Captured · Microsoft's own timestamp · revisionjv6k7mkxOriginal text as we first saw it
On November 3, 2026, between 15:00 UTC and 22:37 UTC, Git operations over HTTPS on GitHub.com were degraded. Clones, fetches, and pushes over HTTPS were slow, and some failed. For about half of Git HTTPS traffic, median response time rose from under 100 milliseconds to as high as 5 seconds. On average, the error rate was 0.65% and peaked at 1.29% of requests to the service. Git operations over SSH were not affected. <br /><br />This was due to a buildup of excess connections on the hosts that serve Git HTTPS traffic, which saturated their network capacity and caused slow and failed responses for all clients served by those hosts. <br /><br />We mitigated the incident by blocking a client traffic pattern at our load balancers and reloading them, which released the excess connections. The block also prevented some users of one third-party integration from performing Git operations over HTTPS until we removed it on November 4 at 13:26 UTC. <br /><br />We are working to alert on sustained Git HTTPS latency sooner and make load balancer mitigations take effect on existing connections immediately.