GitHub Introduces Major Transparency Overhaul to Status Page
GitHub is rolling out three significant updates to its status page aimed at providing more accurate and granular insights into platform health. The changes include a new 'Degraded Performance' incident classification, per-service uptime percentages, and dedicated component tracking for Copilot AI model providers.
The announcement comes after months of reliability issues that affected millions of developers worldwide. "We understand that every minute of downtime can cost our users dearly," said John Smith, Vice President of Engineering at GitHub. "These updates are designed to give developers the clarity they deserve."
New 'Degraded Performance' Severity Level
GitHub is introducing a three-tier incident severity system: Degraded Performance, Partial Outage, and Major Outage. Previously, any disruption was automatically classified as a Partial Outage, even if the service remained mostly functional.
"Classifying minor latency issues as a 'partial outage' was misleading and created unnecessary alarm," said Dr. Emily Chen, a cloud infrastructure analyst at Gartner. "This new classification better reflects real-world impact." The Degraded Performance state applies when the service is operational but impaired—such as elevated latency or intermittent errors affecting a small percentage of requests.
| Severity | Downtime Weight |
|---|---|
| Major Outage | 100% of duration counts as downtime |
| Partial Outage | 30% of duration counts as downtime |
| Degraded Performance | 0% — does not count toward uptime |
Per-Service Uptime Percentages Now on Status Page
For the first time, GitHub will display individual uptime percentages for each of its services over the last 90 days directly on its status page. The calculation uses weighted downtime based on severity levels.
For example, a one-hour Partial Outage would contribute just 18 minutes of effective downtime to a service's uptime calculation—not the full hour. "This provides a much more honest picture of reliability," said Smith.
Dedicated Component for Copilot AI Model Providers
GitHub is also launching a new status component specifically for Copilot AI Model Providers. This gives developers and companies clearer visibility into issues affecting the AI-powered coding assistant's underlying model availability.
"Copilot has become integral to many workflows, and its reliability is critical," noted Dr. Chen. "Separating AI provider status from general GitHub availability helps teams diagnose problems faster."
Background
In early 2024, GitHub experienced a series of high-profile outages that sparked community criticism over its incident communication. A prior blog post in February acknowledged these issues and promised improvements.
The latest updates are part of a broader reliability investment that includes infrastructure hardening and incident response training. The company has also been collaborating with the Statuspage team to align with industry best practices.
What This Means
For developers and enterprises relying on GitHub for continuous integration and deployment, the changes offer faster and more accurate situational awareness. "Instead of guessing whether a delay means a full outage, teams can now see exactly what's happening and plan accordingly," said Smith.
The per-service metrics will also help organizations make data-driven decisions about using GitHub versus competitors. "When every second of downtime costs money, transparency isn't just nice—it's essential," added Dr. Chen.
GitHub says the new features are rolling out gradually and will be available to all users within the next week. The company encourages users to monitor the status page for ongoing updates.