🚨GitHub Outage: 8-Hour Downtime Hits Central US
GitHub's Reliability Issues Push Developers to Look Elsewhere
TL;DR
GitHub experienced an 8-hour outage on Aug 17-18 due to load balancer issues and a retry bug. The incident highlights reliability concerns and may push developers towards alternatives like Cursor or OpenAI.
GitHub faced an unprecedented 8-hour outage on August 17, impacting services across the board. Load balancers in Central US were saturated, causing cascading failures due to misconfigured autoscaling policies and a retry bug in Visual Studio Code. The outage affected Issues, Pull Requests, APIs, Actions, and Copilot, with some services taking over five hours to recover fully. GitHub's reliability issues are not new; this latest incident may be the tipping point for developers seeking alternatives like CloudBees' Cursor or OpenAI.
Key Points
GitHub outage lasted from 13:28 UTC Aug 17 to 21:15 UTC, impacting multiple services.
Load balancers in Central US facility were saturated, causing a cascading failure due to autoscaling policy misconfigurations.
Visual Studio Code's optimistic retry logic overloaded internal load balancers during the outage.
GitHub will correct autoscaling policies and review retry limits post-outage.
CloudBees CEO noted startups like Cursor and OpenAI are building competitive solutions.
Why It Matters
If you're relying on GitHub for critical development workflows, this outage is a wake-up call. Services like Actions and Copilot were down, impacting CI/CD pipelines and code completion tools. The reliability issues highlight the need to evaluate alternatives or diversify your tech stack.
Frequently Asked Questions
Why does this matter?
If you're relying on GitHub for critical development workflows, this outage is a wake-up call. Services like Actions and Copilot were down, impacting CI/CD pipelines and code completion tools. The reliability issues highlight the need to evaluate alternatives or diversify your tech stack.
What happened?
GitHub experienced an 8-hour outage on Aug 17-18 due to load balancer issues and a retry bug. The incident highlights reliability concerns and may push developers towards alternatives like Cursor or OpenAI.
Comments
Be the first to comment
Enjoyed this article?
Get it daily. 7am. Free. Reads in 5 minutes.
Join 3,211 builders reading daily.