2 weeks ago
GitHub outage disrupts web, API traffic and downloads
GitHub is a website where people store and work on computer code together.
On August 17, GitHub started having big problems.
Many of its services stopped working properly.
People had trouble loading pages and using the website's tools.
Downloading code files failed about half the time.
The company said about 20 out of every 100 requests to the website were failing.
Even some security features used by companies were affected.
GitHub's workers were trying to fix the problem.
They did not know yet what caused it.
The problem was still happening when they gave their latest update.
GitHub experienced a widespread service disruption beginning around 1:40 pm UTC on August 17.
Web experiences and API traffic saw an approximately 20% error rate.
Archive downloads and raw repository content downloads saw approximately 50% failure rates.
Affected services included Pull Requests, Issues, Actions, API Requests, Webhooks, Pages, Git Operations and Copilot.
As of the latest update at 16:16 UTC, GitHub had not identified the root cause and was continuing mitigation efforts.
- Who
- GitHub, the software development platform, and its users
- What
- A widespread service disruption affecting multiple GitHub services
- Where
- GitHub's online platform
- When
- Beginning around 1:40 pm UTC on August 17, ongoing as of 16:16 UTC
- Why
- Root cause not yet identified; GitHub was investigating and deploying mitigations
Key facts
- Incident start
- 1:40 pm UTC, August 17
- Web/API error rate
- Approximately 20%
- Download error rate
- Approximately 50%
- Affected services
- Pull Requests, Issues, Actions, API Requests, Webhooks, Pages, Git Operations, Copilot
- Enterprise services affected
- SAML, OIDC authentication, SCIM, Team Sync
- Latest update
- 16:16 UTC
- Root cause
- Not yet identified
- Status
- Unresolved, mitigation ongoing







