Cirrus Runners had 5 outages in the last 2 years totaling 1h 8m of downtime — averaging 0.2 incidents per month.
There were 5 Cirrus Runners outages since December 12, 2025 totaling 1h 8m of downtime. Each is summarised below — incident details, duration, and resolution information.
Recent Sequoia and Tahoe image updates are getting propagated slower then usual which causes delays in job scheduling. The scheduling times are getting improved and should be completely resolved shortly.
Runners operate normally. Dashboard is still not available until DNS records propagate.
The root cause has been detected and we are deploying a fix. Scheduling delays should be resolved shourtly after.
Scheduled Linux runners are not getting picked up
Jobs that got queued due to concurency limit are taking longer to start executing after queue is cleared up.
A fix is deloyed and queues are recovering now
Everything should operate normally