I was testing a cluster, have 4 nodes,
node1 localhost:26251 /data/store1
node2 localhost:26252 /data/store2
node3 localhost:26253 /data/store3
node4 localhost:26254 /data/store4
when the cluster is running, i remove all the files of the node 4 , then restart all nodes.
i found these problems:
Admin UI can’t work normally.
node number from 4 to 5.
and “View nodes list” cat not open.
Fortunately, “cockroach sql” can work on every node, and data is not lost.