This message was deleted.

I think you have to delete the existing etcd db on that node first

that is a setting in your config.yaml that is only relevant on the initial node join. It should be removed during a cluster-reset operation, so you’ll just need to remove the line in your

/etc/rancher/rke2/config.yaml

that is like:

server: <https://aa.bb.cc.dd:9345>

thank you @gray-lawyer-73831, it works.