GitHub Faces Major Outage as Actions Service Struggles, Accelerates Move to Azure

On Thursday, August 6, 2026, GitHub’s Actions service suffered a prolonged outage that disrupted the Web UI, REST and GraphQL APIs, Packages, Copilot, and Git operations across Enterprise Cloud environments. The incident, lasting over seven hours, triggered widespread 5xx errors and caused login failures, queued or rejected workflow runs, and unavailable package registry endpoints. GitHub’s internal monitoring detected the spike in errors within 19 minutes, and an automated infrastructure update that altered a runtime configuration value on running virtual machines—normally protected by a safeguard—disrupted service discovery and left traffic routers without backends. The company restored correct configuration values to mitigate the issue, but the cascading failures delayed recovery. In response, GitHub is accelerating its architectural roadmap to isolate, increase resiliency, and scale the Actions service by moving it to Azure, where it can absorb traffic spikes. July’s progress saw more than half of monolith read traffic shift to Azure Central US, authentication tables migrate to dedicated infrastructure, and repository content traffic run fully from Central US, reducing shared failure points. GitHub now targets 70% of read traffic and 30% of write traffic in Central US by the end of the quarter, aiming to eliminate datacenter dependence and improve overall availability.