Hi,
Am trying to add a server to a group of 4 clusters on different servers and one fails with an error when i attempt to add it as follows
[sercom@smscappsvr02 bin]$ ./redis-trib.rb create 10.161.1.150:6379 10.161.1.149:6379 10.161.1.154
>>> Creating cluster
[ERR] Node 10.161.1.149:6379 is not empty. Either the node already knows other nodes (check with CLUSTER NODES) or contains some key in database 0.
on the node that it is complaining about if i run cluster info from its redis-cli prompt i get the following output:
10.161.1.149:6379> cluster info
cluster_state:fail
cluster_slots_assigned:16384
cluster_slots_ok:8243
cluster_slots_pfail:8141
cluster_slots_fail:0
cluster_known_nodes:2
cluster_size:2
cluster_current_epoch:1
cluster_my_epoch:0
cluster_stats_messages_sent:2890
cluster_stats_messages_received:0
10.161.1.149:6379>
Is there a way i can reset that node or fix the error as it seems the node unsuccessfully tried to join a cluster previously? NB: I think i may have tried to use the default create-cluster on that node i honestly cant remember. Is there a way that i can reverse its effects?
Comment From: michnus
shutdown per redis server and remove node-6379-conf file producted last time you run cluster create scripts. one important thing is that make sure the port 16379 which every nodes in cluster conmunication with each other is opened
Comment From: qiyue49
1 remove the aof/rdb backup file 2 remove the node_conf file 3 flushdb if necessary
Comment From: sajanmani
@qiyue49 - your comment helped me fix the same issue reported - took me hours to figure out. Thanks a ton. SM
Comment From: GypsyCosmonaut
Also, if you're running multiple servers on one machine on different ports, make sure, they have different rdb and log files