Since the update to a newer CI image which of course brings in a much newer kernel I keep seeing CI agent stopped responding! errors from cirrus there.
Possible reasons is that something is killing the agent process, maybe we run out of memory somehow? OR maybe something else breaks the networking?
It is a flake and passes on rerun so whatever it is, it is not consistent.
The main issue we do not get any useful logs either due that so not sure how we best debug this
Two example cirrus tasks:
https://cirrus-ci.com/task/5416603430420480
https://cirrus-ci.com/task/5638759170965504
I think I only saw this on rawhide so far so it seems safe to assume it is something in the image which causes it and not cirrus itself.
Since the update to a newer CI image which of course brings in a much newer kernel I keep seeing
CI agent stopped responding!errors from cirrus there.Possible reasons is that something is killing the agent process, maybe we run out of memory somehow? OR maybe something else breaks the networking?
It is a flake and passes on rerun so whatever it is, it is not consistent.
The main issue we do not get any useful logs either due that so not sure how we best debug this
Two example cirrus tasks:
https://cirrus-ci.com/task/5416603430420480
https://cirrus-ci.com/task/5638759170965504
I think I only saw this on rawhide so far so it seems safe to assume it is something in the image which causes it and not cirrus itself.