User Tools

Site Tools




2016-11-20 17:51:37.412 [39620] main/104/applier/localhost:11000 I> can't read row
2016-11-20 17:51:37.412 [39620] main/104/applier/localhost:11000 E> ER_CLUSTER_ID_MISMATCH: Cluster id of the replica a438c8ee-d8cf-4e11-9725-0bd11368668a doesn't match cluster id of the master 222d9d8f-26f6-4cc2-9003-b7b6821aa205

I got this when I did box.cfg{replication_source = user:password@some_host:11000}

I did os.exit(), deleted all .snap and .xlog files, started again - and it's running well. What happened? (2016)

A. You need to tell a replica who its master is ahead of time, before it is bootstrapped. Otherwise, it assumes it's the master, and, as a result, you get two replica sets, not one.

When starting a replica, set replication_source to the master’s URI before bootstrapping.


Enter your comment. Wiki syntax is allowed:
replication/replication_source.txt · Last modified: 2018/05/25 21:39 by eabates

Real Time Web Analytics