Incident BriefingsGitHub August 2026 load-balancer outage

GitHub August 2026 load-balancer outage

GitHub reported an availability failure involving load balancers and retries. The provider’s account does not establish a data breach.

Last modified

Summary

  • Environment: GitHub.com service infrastructure
  • Operational impact: Errors and latency across multiple services
  • Financial impact: No quantified financial impact in the incident report
  • Record status: Documented record. Reviewed October 7, 2026; updated as evidence emerges.

What happened

GitHub reported an availability incident on August 17, 2026, spanning 13:28–21:15 UTC. Its incident report describes infrastructure saturation and recovery. [1]

Impact

  • Errors and latency affected Issues, pull requests, APIs, Actions and Copilot. [1]
  • This record does not establish unauthorized access or data exposure.

Timeline

  1. Service disruption and recovery

    The provider-reported interval ran from 13:28 to 21:15 UTC; it does not mean every service was unavailable throughout.

    [1]

  2. Briefing updated

    This briefing was last reviewed and updated on October 7, 2026.

Threat Group & Attack Vector

GitHub attributed the failure to Central US load-balancer saturation, a sidecar autoscaling misconfiguration and exhausted HAProxy flow limits. Retry amplification delayed Copilot authentication recovery. [1]

Actors

  • No actor is attributed to this operational failure in the canonical record.

TTPs

  • No MITRE ATT&CK technique is mapped to this availability event.

Response

GitHub rerouted traffic, paused affected HAProxy nodes and reduced retries to restore service. [1]