I’ve used rke (1.2.7) to remove old (broken) etcd nodes. They are gone and the kube-apiserver on the control plane “appears” to have a new

--etcd-servers

command-line argument with the functional etcd machines but

kubectl get no

still shows the old etcd nodes! At the moment there is only one etcd node: -

$ docker exec 04 etcdctl member list -w table
+------------------+---------+---------------------+-----------------------------+-----------------------------+------------+
|        ID        | STATUS  |        NAME         |         PEER ADDRS          |        CLIENT ADDRS         | IS LEARNER |
+------------------+---------+---------------------+-----------------------------+-----------------------------+------------+
| 9d7b6280a957bef4 | started | etcd-192.168.253.47 | <https://192.168.253.47:2380> | <https://192.168.253.47:2379> |      false |
+------------------+---------+---------------------+-----------------------------+-----------------------------+------------+