Recovering an unavailable ETCD cluster node

Before proceeding with recovery, install module etcd-helper.

If the IP address and name of the API server have not changed, but the data on the Control Plane node is corrupted, recreate the node in the ETCD cluster by running the etcdhelper rejoin command.

/usr/local/bin/etcdhelper rejoin

Please note! If the ETCD container on the node is in Running status, then the etcdhelper rejoin command cannot be executed.

As a result of executing the command, the following will happen:

• Removing a non-working node from the ETCD cluster.
• Stop kubelet.
• Removing the ETCD container.
• Renaming the directory with ETCD data.
• Adding a new node to the ETCD cluster.
• Launch kubelet.