Rook and ceph on kubernetes

Question

I am new to Kubernetes. I am in need of integrating rook and ceph, adding NFS as block storage. Does anyone have any working examples? I followed https://www.digitalocean.com/community/tutorials/how-to-set-up-a-ceph-cluster-within-kubernetes-using-rook this document and I am getting errors(stuck at container creating, stuck at pod initializing) while creating ceph cluster in rook on Kubernetes. Any help would be appreciated.

kubectl get pod -n rook-ceph  

NAME                                            READY   STATUS              RESTARTS   AGE
csi-cephfsplugin-provisioner-5bcd46f965-42f9r   0/5     ContainerCreating   0          12m
csi-cephfsplugin-provisioner-5bcd46f965-zszwz   5/5     Running             0          12m
csi-cephfsplugin-xcswb                          3/3     Running             0          12m
csi-cephfsplugin-zwl9x                          3/3     Running             0          12m
csi-rbdplugin-4mh9x                             3/3     Running             0          12m
csi-rbdplugin-nlcjr                             3/3     Running             0          12m
csi-rbdplugin-provisioner-6658cf554c-4xx9f      6/6     Running             0          12m
csi-rbdplugin-provisioner-6658cf554c-62xc2      0/6     ContainerCreating   0          12m
rook-ceph-detect-version-bwcmp                  0/1     Init:0/1            0          9m18s
rook-ceph-operator-5dc456cdb6-n4tgm             1/1     Running             0          13m
rook-discover-l2r27                             1/1     Running             0          13m
rook-discover-rxkv4                             0/1     ContainerCreating   0          13m

Answer 1

It may be also the case that one of your nodes is in bad state, as it seems that some pod replicas are failing to start. You can confirm by running

kubectl get pod -o wide | grep -v Running

Possible all failing pods are running on the same node. If that is the case you can inspect the problematic node with

kubectl describe node [node]

Answer 2

Use kubectl describe pod <name> -n rook-ceph to see the list of events, it is on the bottom of the output. This will show where the pods get stuck.