GitHub detailed a nearly eight-hour outage that disrupted Issues, Pull Requests, APIs, Actions, and Copilot. The incident stemmed from saturated load balancers and related infrastructure problems in its Central US facility. A misconfigured autoscaling policy failed to account for an Istio sidecar’s concurrency limit, allowing a cascading failure. Optimistic retry logic and a latent Visual Studio Code bug drove huge internal traffic spikes, slowing recovery. Engineers reduced gateway retries, adjusted load balancer behavior, and plan to fix autoscaling, retry limits, Istio settings, and VS Code behavior. Ongoing instability could push developers toward competitors, including Cursor, OpenAI, and smaller startups.
This update represents a notable development in the Board sector. Organizations and founders tracking this space should evaluate potential strategic and technical implications on their operations.