On June 6, 2026 between 16:18 UTC and 17:01 UTC, users experienced elevated error rates when performing Git operations (cloning, fetching, downloading archives) and accessing package registries. The issue affected users whose traffic was routed through our European infrastructure.
During this time, on average 0.95% of Codeload requests and 9.2% of Package Registry requests failed with server errors. At peak, the Codeload error rate reached 1.76% and Package Registry errors reached 27%.
The root cause was a planned network circuit migration that disrupted connectivity at one of our points of presence. Our process for shifting traffic away from the site did not operate as expected, resulting in a small amount of production traffic to continue being serviced at the effected site during the maintenance window. The issue was mitigated by rolling back the network change, restoring normal connectivity. Services fully recovered by 17:01 UTC.
To reduce the likelihood of similar incidents in the future, we are reviewing our site drain process to make it more verbose and add visibility so any unexpected behavior is caught earlier.
Posted Jun 06, 2026 - 17:07 UTC
Update
Packages is experiencing degraded performance. We are continuing to investigate.
Posted Jun 06, 2026 - 16:56 UTC
Investigating
We are investigating reports of impacted performance for some GitHub services.