Shresth Suman:
Hi everyone,
I was trying to setup a kubernetes cluster with CRI-O as container runtime. Everything works well with version 1.20. However, with version 1.21, kubeadm init command fails. If anyone has any idea, feel free to share it.
More version details -
### Working version details -
OS=xUbuntu_20.04
VERSION=1.20 #cri-o version
apt install -qq -y kubeadm=1.20.5-00 kubelet=1.20.5-00 kubectl=1.20.5-00
### Not working version details
OS=xUbuntu_20.04
VERSION=1.21 #cri-o version
apt install -qq -y kubeadm=1.21.0-00 kubelet=1.21.0-00 kubectl=1.21.0-00
Daz Mac:
What is the error? I found i needed ubuntu 20.10 for 1.21
Also i found cgroup manager issue it defaults to systemd in 1.21 for kubelet and i was configuring it as cgroupfs (like i did in 1.20)
Shresth Suman:
@Daz Mac pfa the ss error with version 1.21… I used ubuntu 20.04
and the following cgroup config (https://kubernetes.io/docs/setup/production-environment/container-runtimes/#cgroup-driver)-
cat >>/etc/crio/crio.conf.d/02-cgroup-manager.conf<<EOF
[crio.runtime]
conmon_cgroup = "pod"
cgroup_manager = "cgroupfs"
EOF
Daz Mac:
@Shresth Suman that is exactly what i was doing creating that file. I removed it in doing a 1.21 build as kubelet defaults to systemd.
https://kubernetes.io/docs/tasks/administer-cluster/kubeadm/configure-cgroup-driver/#configuring-the-container-runtime-cgroup-driver|https://kubernetes.io/docs/tasks/administer-cluster/kubeadm/configure-cgroup-driver/#configuring-the-container-runtime-cgroup-driver
I can’t 100% remember what specifically the error message i got and i’m away from my laptop today. However i am pretty sure it was the same as your first image with the kubeadm failing.
Shresth Suman:
@Daz Mac Thanks for the info… Resolved the issue now… thanks!!