`Pod for etcd not synced (pod sandbox not found), ...
# rke2
b
Pod for etcd not synced (pod sandbox not found), retrying
I already have ufw and apparmor masked (have for years on this cluster) and I don't get any other relevant logs
c
check the kubelet.log for errors
b
@creamy-pencil-82913 I tried to manually launch the kubelet command spewed to the console..
unknown flag: --alsologtostderr
c
don’t do that
just look at the logs
b
looks like its failing due to cgroupv1
c
Did you read the release notes before upgrading? https://github.com/kubernetes/kubernetes/blob/master/CHANGELOG/CHANGELOG-1.35.md#deprecation > • ACTION REQUIRED: `failCgroupV1`will be set to true from 1.35. This means that nodes will not start on a cgroup v1 by default. This puts cgroup v1 into a deprecated state. (#134298,@kannon92) >
b
cool, I will work on getting the nodes updated
thanks for the pointers
c
I would have appreciated a little (*) note for OS-es that default to cgroupv1 in the support matrix of k8s 1.35+. Like RHEL-clone version 8.x. That would have saved me some sweat yesterday. Yes I should have read the k8s release notes, but when you have such a nice compatibility list this warrants a warning IMO.
Also, is there a downside to enabling cgroupv2 in rhel8 type OS instead of allowing v1 on k8s?