Delays starting Actions runs

[investigating] We are investigating reports of degraded performance for Actions

[investigating] Approximately 5% of GitHub Actions runs on GitHub-hosted runners are experiencing run start delays exceeding 5 minutes. A small portion of these runs may fail after extended delays. We have identified the cause and are working on a mitigation.

[investigating] We are continuing to work on a mitigation.

[investigating] Actions is experiencing degraded availability. We are continuing to investigate.

[investigating] Approximately 30% of GitHub Actions runs on GitHub-hosted runners are experiencing run start delays exceeding 5 minutes. A smaller percentage of those are exhausting retries and failing to start.

[investigating] Approximately 30% of GitHub Actions runs on GitHub-hosted runners are experiencing run start delays exceeding 5 minutes. A smaller percentage of those are exhausting retries and failing to start.<br />This has caused some customers to exceed their hosted compute concurrency and experience increased impact.<br /><br />We are continuing to working on infrastructure mitigations. <br /><br />Next update in one hour.

[investigating] Pages is experiencing degraded performance. We are continuing to investigate.

[investigating] Actions is experiencing degraded availability. We are continuing to investigate.

[investigating] We're seeing Actions and Pages recovery. <br /><br />For a period of approximate 20 minutes ~96% of GitHub Actions runs on GitHub-hosted runners were failing to start, but has now recovered and we are seeing jobs processing.<br /><br />GitHub pages builds were also failing during that period, but Pages are still accessible.<br /><br />We are continuing to monitor for full recovery.

[investigating] Actions is experiencing degraded availability. We are continuing to investigate.

[investigating] We are continuing to monitor slow recovery in Actions and Pages builds as the system works through the high volume of backlog.<br /><br />Customers may see a small rate of  API and job failures as the system is recovering.<br /><br />Copilot Cloud Agent and Copilot Code Review also failed to start for approximately 30 minutes during this incident, and we are monitoring recovery.<br /><br />Pages were accessible throughout the incident.

[monitoring] The degradation affecting Actions and Pages has been mitigated. We are monitoring to ensure stability.

[monitoring] Actions, Pages builds, Copilot Cloud Agent, and Copilot Code review have all recovered and are mitigated.<br /><br />We are continuing to monitor to ensure full recovery, and investigating the health of the affected infrastructure.

[resolved] This incident has been resolved. Thank you for your patience and understanding as we addressed this issue. A detailed root cause analysis will be shared as soon as it is available.

[resolved] On July 9, 2026, between 03:29 UTC and 13:39 UTC, GitHub Actions experienced delayed and failed job starts on GitHub-hosted runners. The incident was caused by an unhealthy state in a backend data service responsible for provisioning hosted runners, preventing runner acquisition for a subset of workloads. During most of the incident, approximately 8% of hosted-runner jobs experienced start delays exceeding 5 minutes, while roughly 2% failed to start.<br /><br />At 13:39 UTC, we restored the health of the backend data replication system, allowing provisioning to recover and the accumulated workflow backlog to drain. Service performance then returned to expected levels. We are improving provisioning-service resiliency, workload distribution, and capacity balancing to reduce the likelihood and impact of similar incidents.

[resolved] On July 9, 2026, between 03:29 UTC and 13:39 UTC, GitHub Actions experienced delayed and failed job starts on GitHub-hosted runners. The incident was caused by an unhealthy state in a backend data service responsible for provisioning hosted runners, preventing runner acquisition for a subset of workloads. During most of the incident, approximately 8% of workflow runs on hosted runners were delayed by more than 5 minutes, while roughly 2% failed to start.<br /><br />At 13:39 UTC, we restored the health of the backend data replication system, allowing provisioning to recover and the accumulated workflow backlog to drain. Service performance then returned to expected levels. We are improving provisioning-service resiliency, workload distribution, and capacity balancing to reduce the likelihood and impact of similar incidents.