Current Status

ongoing.

Error looks like

Anything can fail

Incident timeline (all times pacific)

Include when the incident began, when it was detected, mitigated, root caused, and finally closed.

User impact

How does this affect users of PyTorch CI?

Root cause

https://www.githubstatus.com/incidents/smf24rvl67v9 and then https://www.githubstatus.com/incidents/lcw3tg2f6zsd

Mitigation

How did we mitigate the issue?

Prevention/followups

How do we prevent issues like this in the future?