Service Alert: Performance Degradation for GitHub Pages and Actions

Status: Degraded Availability $\rightarrow$ Status: Resolved

🔔 Stay Informed: Subscription Options

Users can opt-in to receive real-time notifications regarding incident creation, updates, and resolutions through several channels:

1. Email Notifications

• Requirement: Provide a valid email address.
• Verification: A One-Time Password (OTP) is required for security.
• Support: Options to resend the OTP if not received.

2. SMS Alerts

Notifications can be sent via text message to a global list of regions.

• Coverage: Extensive international support including, but not limited to:
  • North America: USA (+1), Canada (+1)
  • Europe: UK (+44), Germany (+49), France (+33), Italy (+39)
  • Asia/Pacific: India (+91), Japan (+81), Australia (+61), China (+86)
  • Africa/Middle East: Egypt (+20), Nigeria (+234), UAE (+971)
  • (And many others across Afghanistan, Albania, Algeria, etc.)
• Process: Select country code $\rightarrow$ Enter phone number $\rightarrow$ Verify via OTP.

3. Slack Integration

Integrate status updates directly into your workspace via the "Subscribe via Slack" option.

4. Webhook Integration

For automated system responses, configure a webhook:

• Webhook URL: The destination endpoint for JSON payloads.
• Fallback Email: An address to notify you if the endpoint fails.

[!IMPORTANT] Legal & Privacy: By subscribing, you agree to the GitHub Privacy Policy, the Atlassian Terms of Service, and the Atlassian Privacy Policy. All forms are protected by reCAPTCHA; therefore, Google's Privacy Policy and Terms of Service apply.

---

🛠 Incident Report: GitHub Actions

Executive Summary

"This incident has been resolved. We appreciate your patience and understanding while we worked to fix this issue. A comprehensive root cause analysis (RCA) will be published as soon as it is finalized."

Incident Timeline (August 07, 2026)

The duration of the event can be represented as $\Delta t = T_{resolved} - T_{start}$.

Time (UTC)	Status	Update Details
00:05	Update	System queues have been cleared; new jobs are processing normally.
00:06	Update	Mitigation applied to both Actions and Pages.
00:59	Monitoring	The service degradation has been successfully mitigated.
02:03	Update	Investigating reports of slow recovery for Actions Runner Controller (ARC) runners.
02:04	Resolved	Incident officially closed.

Technical Breakdown & Recovery

The incident caused specific issues with the Actions Runner Controller (ARC), where certain runner pods became trapped in an idle state.

Recovery Flow

Manual Remediation

If you are experiencing stuck pods, you may use the following command:

# Example: Deleting a stuck ARC runner pod
kubectl delete pod <pod-name> -n <namespace>

Future Prevention

GitHub has committed to implementing an automatic recovery mechanism in upcoming releases of both the Actions Runner and the Actions Runner Controller to eliminate the need for manual pod deletion.

⚠️ Data Loss Warning

Certain events that trigger workflows (such as push or pull_request) were not processed during the outage. These events cannot be replayed by the system.

Required User Actions:

• Push a new commit to the branch.
• Update the existing pull request to trigger a refresh.
• Manually trigger the workflow via the GitHub UI (where applicable).