The correct way to centrally maintain PVE clusters

[sky_me]

hi
I'm confused and I don't feel like I'm going the right way. When a host in my cluster has a hardware failure and needs to be shut down for replacement. I have to do:
1. Migrate the virtual machine on the host machine.
2. Click the shutdown button on the host interface on the web
3. After shutting down, replace and turn on the
4. Check the status of the host in the cluster
5. Migrate virtual machines back in batches.
But I also read that according to this method, the host machine should be separated from the cluster before proceeding.

#systemctl stop pve-cluster.service
#systemctl stop corosync.service
#pmxcfs -l
#rm /etc/pve/corosync.conf
#rm -rf /etc/corosync/*
#killall pmxcfs#pvecm delnode <faulty_node_name>
After the repair is completed, rejoin the cluster through the web UI.
What is the correct one? I'm confused.

thks

 

[LnxBil]

But I also read that according to this method, the host machine should be separated from the cluster before proceeding.

Where have you read this?

 

[sky_me]

hh

Where have you read this

Where have you read this?

Hi LnxBil
I was looking for general blogs online and didn’t see the official operation manual.
https://www.cnblogs.com/jackadam/p/15763362.html

 

[LnxBil]

[...] and didn’t see the official operation manual.

It is here.