GitLab Outage
GitLab experienced a 502 Bad Gateway error, impacting all users. The issue is currently unresolved.
since
Here's why. Incidents from Thursday, September 24, 2026 (UTC), ranked byBreakage Velocity.What is Breakage Velocity?A live score of how fast an outage or regression is spreading. The more people hit it, and the faster, the higher it climbs; it cools off as things calm down.Open-ended scale: under 25 low · 25+ elevated · 100+ high · 300+ extreme.
15 outages, 3 still active today. Worst: GitLab (55 d 5 h).
GitLab experienced a 502 Bad Gateway error, impacting all users. The issue is currently unresolved.
since
The OpenAI health website's API rates were elevated, affecting its availability and response times. No specific version or configuration is mentioned.
since
Around 23:31 UTC on 2026-09-23, the Key Value services created prior to 2024-03-05 were inadvertently restarted due to a configuration change. No further restarts are planned or expected.
– UTC
Elevated error rates on the OpenAI GPT-6 Astra Pro service have been resolved. No affected versions provided.
– UTC
A technical incident caused errors accessing Netlify-hosted sites, affecting users. The issue has been resolved with no longer elevated error rates.
– UTC
Database issues in the GRU (Sao Paulo, Brazil) region have been resolved. No affected versions provided.
– UTC
An issue with project lifecycle management in the eu-west-1 region has been resolved. No affected versions or workarounds are provided.
– UTC
A feature on Heroku experienced degradation, affecting all users. The issue has been resolved.
– UTC
Private networking functionality affected customer applications. Incident resolved.
– UTC
Permission errors in the Supabase Dashboard have been resolved. No further permission-related issues were observed after the fix.
– UTC
Storage search functionality is failing for projects that have been restored. The issue has been resolved and hardening measures are being implemented to prevent recurrence.
– UTC
Some clients experienced connection timeouts to their Redis databases hosted in the fra or gig regions due to a planned rolling upgrade where replicas were left out of service and returned to Fly routing before they were ready. The issue was resolved by 13:05 UTC.
– UTC