Percona Cluster Functionality

Have existing percona cluster and new percona cluster. Both have a few slight configuration differences. Would like to know functionality wise, what the difference is, if any.

Cluster A - 3 nodes , my.cnf file server_id is respectively 1, 2 and 3 . does server id have any effect on if servers are rebooted?

Cluster B - 3 nodes +DR asynchronous replication host. Server_id for 3 nodes is the same. and server_id is different than the other 3 nodes on DR node. how do reboots work in this scenario since these server nodes are same server_id’s? If no different that’s fine. Just trying to clarify if there is any difference between 2 clusters in this regard.

1 Like

Hi @jasonfe33 ,
server_id sticks for the life of the server until you change that parameter. Reboot away, the value stays.

As long as cluster A never replicates to cluster B, you can have duplicate server_ids.

cluster A - server_id 1,2,3
cluster B - server_id 1,2,3+4(async)

There are no issue there because A never talks to B and B never talks to A. There would only be an issue if A replicated to B.

1 Like