We have drained the runners of stale workloads, prevented the issue from occurring in the future, and seen full recovery across the fleet of Windows Runners.

We are seeing recovery of some workloads now that a fix has been pushed and some stale workloads were removed

Windows Runners have been degraded and windows sandboxes cannot be created. Some existing workloads are experiencing errors due to the degradation.

New creations will fail at this time.

We are investigating the underlying cause of the issue