Closed
Description
What happened?
When using crun as the runtime, in some cases a pod's graceful shutdown isn't respected as the node is shutting down. The CRI-O team is investigating, but this issue is to warn users
What did you expect to happen?
a pod is given it's full graceful shutdown period before terminated and the node is rebooted
How can we reproduce it (as minimally and precisely as possible)?
Not totally clear yet, we've mostly seen it in CI now.
Anything else we need to know?
No response
CRI-O and Kubernetes version
$ crio --version
# paste output here$ kubectl version --output=json
# paste output hereOS version
# On Linux:
$ cat /etc/os-release
# paste output here
$ uname -a
# paste output here