--- title: "The August 17 outage, and the work ahead" url: https://daily.dev/posts/the-august-17-outage-and-the-work-ahead-5itrcdatl source_url: https://github.blog/news-insights/company-news/the-august-17-outage-and-the-work-ahead type: article source: "GitHub Blog" published: 2026-08-20T18:39:02.944Z updated: 2026-08-20T21:46:34.459Z tags: ["github", "azure", "github-actions"] reading_time: 4 upvotes: 0 comments: 0 language: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # The August 17 outage, and the work ahead **[GitHub Blog](https://daily.dev/sources/ghblog)** · 4 min read · 0 upvotes · 0 comments ## Summary GitHub's CTO details the root cause of the August 17 outage that lasted nearly 8 hours, disrupting github.com, authentication, Actions, APIs, pull requests, issues, and Copilot. The failure stemmed from a critical infrastructure component in a Central US data center that failed to scale under a new traffic peak, compounded by a client-side retry loop in Copilot services during recovery. Neither this incident nor the August 6 Actions failure was caused by a code change; both were capacity failures driven by commit volume growing from 1.4 billion to 2.9 billion monthly since April. GitHub has added over 3 million CPU cores, 120 petabytes of storage, and shifted 58% of platform load to Azure (up from 12% in May). Going forward, the company is applying consistent retry limits and budgets, reviewing alert thresholds, and isolating critical systems to reduce shared dependencies. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What caused the GitHub outage on August 17 A critical infrastructure component in GitHub's Central US data center failed to scale when traffic reached a new peak, causing capacity pressure that led to authentication failures across multiple services. The outage lasted 7 hours and 47 minutes and affected github.com, authentication, GitHub Actions, APIs, pull requests, issues, and Copilot. Recovery was delayed further by a client-side retry loop in Copilot services that increased traffic during the recovery phase. _daily.dev helps engineers track incident postmortems like this one to inform their own reliability practices._ ### How much of GitHub's platform load runs on Azure now Azure serves roughly 58% of GitHub's platform load and half of all Git operations, up sharply from just 12% of platform load in May. GitHub accelerated this migration alongside adding more than 3 million CPU cores and 120 petabytes of high-speed storage as part of its post-outage reliability push following capacity failures in August. _engineers evaluating cloud migration strategy can follow platform infrastructure shifts like this on daily.dev._ ### How is GitHub preventing retry storms after its 2026 outages GitHub is applying consistent retry limits, retry budgets, and variable timeouts across service-to-service interactions to prevent retry storms and cascading load. This change followed the August 6 and August 17 incidents, in which a client-side retry loop in Copilot services worsened traffic during recovery and delayed restoring service. _teams designing resilient retry logic can compare real-world incident lessons like this on daily.dev._ --- Tags: [#github](https://daily.dev/tags/github), [#azure](https://daily.dev/tags/azure), [#github-actions](https://daily.dev/tags/github-actions) [View this post on daily.dev](https://daily.dev/posts/the-august-17-outage-and-the-work-ahead-5itrcdatl)