After Node Reboot, node's Trident Pod work slow

[sjlee-tech]

Hello everyone

Below informantion is Our site environment K8S : 1.19.16 Trident : Server 21.07.1 / Client 20.10.0 Ubuntu 20.04 LTS

We did some node reboot, by our schedule after reboot node,

node's trident pod state was 1/2 running , so our POD cannot mount the storage After a very long time, the pod starts operating normally and is normally mounted on the storage. Even if many nodes are rebooted at the same time, the problem only occurs in random nodes. we can't solve this problem There is no applied network policy. Firewall is also turned off

I created a test environment and tested it. When the node reboots, the same phenomenon as operating in the problem environment is shown. However, in my case, the daemon pod comes up quickly.

Please tell me what could be causing the daemon pods to be slow

after long long long time (about 9hour , 10hour ) csi.sock will make or not so we do reboot node again then, that work more quickly.

What should I check , I want to know what to do