After rook-ceph cluster deployment is complete, the new host discovers all nodes can not be mounted container storage space:
Rancher platform launch container error display:
Deployment does not have minimum availability
Solution is as follows:
I rke deployment k8s cluster, you need to modify rke cluster.yaml file, the plugin directory into the container, add the following:
kubelet:
extra_args:
cgroup-driver: 'systemd'
kubelet:
extra_args:
volume-plugin-dir: /usr/libexec/kubernetes/kubelet-plugins/volume/exec
extra_binds:
- /usr/libexec/kubernetes/kubelet-plugins/volume/exec:/usr/libexec/kubernetes/kubelet-plugins/volume/exec
Update the cluster configuration
cd /etc/rke/
[root@cnvs-kubm-101-103 rke]# ll
total 10716
-rw-r----- 1 root root 121688 Sep 4 11:09 cluster.rkestate
-rw-r--r-- 1 root root 1567 Sep 4 11:00 cluster.yml
-rw-r----- 1 root root 5431 Sep 4 11:08 kube_config_cluster.yml
#执行
rke up
After the cluster upgrade is complete, re-establishment of storage space, the service can start again normally mounted disk.
Reference documents:
https://github.com/rook/rook/blob/master/Documentation/flexvolume.md