Tailscale outage history

Users worldwide faced login and authentication problems beginning around 1:00 PM UTC, with reports from Mumbai, Turin, Washington, DC, and Jacksonville illustrating sign-in and SSO failures. The disruption affected Funnel and Serve domains by blocking new certificate issuance for hostnames, while existing certificates remained valid. Earlier in the day, Funnel showed degradation and certificate-creation issues began to appear, signaling broader impact on authentication workflows rather than a full platform outage. A sequence of fixes and monitoring followed as engineers worked to restore services.

Impact began around 7:00 AM UTC as Tailscale reported coordination server issues that led to authentication failures for many users. Admin console access and the core coordination service were affected, with login problems reported from locations worldwide including the United States, India, Italy, and Thailand. At the same time, Funnel performance degraded and new certificate issuance for Funnel and Serve domains was disrupted, with existing certificates remaining valid for now.

Engineers rolled out a sequence of fixes through the day: the first fix by 8:28 AM UTC; a follow-up fix by 10:47 AM UTC; and another stabilization effort by 4:28 PM UTC that restored Funnel and Certificates operations. As monitoring continued, some components briefly reported renewed issues, but overall the situation improved. By 1:32 AM UTC the Admin console and Coordination service were back up and the service returned to normal operation. The incident affected a broad user base across regions, with a noticeable uptick in authentication-related reports during the afternoon of 2026-08-03 before stabilizing.

Starting on Aug 3, 2026 7:03 AM UTC, Tailscale flagged coordination server issues that led to elevated errors in the coordination service and degraded user experiences. Users across regions reported login/authentication failures, and Funnel performance suffered with timeouts while issuing new certificates for Funnel and Serve domains was affected. Admin tooling and log access were intermittently impacted, though live logs remained accessible via API and the Admin Console during parts of the incident. The impact rippled through multiple features that rely on coordination, including authentication workflows and deployment workflows via Funnel.

Engineers rolled out a series of mitigations through the day, with an initial fix deployed by Aug 3, 2026 8:28 AM UTC and continued monitoring as issues persisted with higher error rates and certificate-issuance problems. Improvements followed in phases: additional fixes deployed around Aug 3, 2026 10:47 AM UTC and again later as coordination and funnel components stabilized. By Aug 4, 2026 1:32 AM UTC, Admin Console and Coordination service were back up, and most user-facing functionality had recovered while some lingering delays remained in earlier-degraded paths.

Some users experienced HTTP 500 errors when attempting to sign in to Tailscale, impacting access to the Admin console and Coordination service for a subset of users. The issue appears to have started with signs of sign-in problems around 4:30 PM UTC.

The status page noted degraded sign-in starting at 5:18 PM UTC, affecting Admin console (console.tailscale.com) and the Coordination service. A fix was deployed at 5:33 PM UTC and monitoring continued, with full recovery by 5:48 PM UTC.

Aperture access for self-serve users was unavailable starting at 10:57 AM UTC and persisted for about one hour and 40 minutes until the service was restored at 12:37 PM UTC. Tailscale engineers investigated the issue and confirmed the problem on the status page, implementing a fix in the period between 11:13 AM UTC and 11:52 AM UTC. The fix was deployed and monitoring continued, with Aperture returning to normal by 12:37 PM UTC.

An outage affected Tailscale's advertising services for the CLI and Admin Console, impacting ability to advertise services and access the Admin Console (login.tailscale.com). The issue began the evening of 9:23 PM UTC and persisted for roughly an hour as engineers rolled out a fix and monitored the service. A fix was deployed by 9:58 PM UTC, with an update at 10:08 PM UTC indicating the fix was in place and the Admin Console was back to normal; the overall service was restored by 10:23 PM UTC.

Users experienced degraded ControlD integration affecting profile selection and device attribution on Tailscale. This led to issues with applying ControlD profile changes and accurate device attribution for some users. The impact began at 1:47 PM UTC and centered on the Coordination service, with ongoing concerns noted as the issue circulated through status updates.

Status updates from Tailscale outline the response: at 1:52 PM UTC the problem was identified and a fix began rolling out. A subsequent update at 3:38 PM UTC reiterated the ongoing resolution, and by 3:47 PM UTC a fix had been deployed with monitoring ongoing. The Coordination service returned to the up state at 4:27 PM UTC, signaling recovery from the incident.

Some users experienced unresponsive logins to the Tailscale Admin Console (login.tailscale.com). The incident began at 10:32 PM UTC and affected admin access for about twenty minutes. At 10:42 PM UTC engineers identified the problem and began rolling out a fix; the Admin Console returned to normal by 10:53 PM UTC.

Status updates described the degraded login state for the Admin Console and then confirmed that a fix had been rolled out and access was restored, with the component returning to healthy by the end of the window.

Tailscale reported a problem in the coordination service that began at 2:43 PM UTC, resulting in an elevated rate of errors affecting coordination-related tasks. The incident led to noticeable impact for users relying on cross-node coordination during the early part of the window. A fix was deployed at 3:23 PM UTC as engineers worked to restore normal operation and monitored for any further issues. The service returned to normal by 6:48 PM UTC.

Users attempting to manage or install Tailscale faced outages starting at 8:17 AM UTC when the Admin console (login.tailscale.com) and the package server (pkgs.tailscale.com) became unavailable due to issues at Amazon CloudFront causing 504 gateway errors. Admin console access and package fetches were impacted, disrupting setup and administration for many customers worldwide. Admin console partial recovery began by 9:07 AM UTC with the Admin console moving from down to up, though pkgs.tailscale.com remained unreachable. The team rolled out a fix for the admin console and by 10:33 AM UTC both services were reported as operational again, with the status gradually stabilizing. A brief period of degraded performance followed, and by 1:08 PM UTC the system was fully up and stable. The incident appears to be tied to CloudFront issues at AWS; the Tailscale team rolled out fixes and continued monitoring for any further degradation.