I was looking into whether I could configure a master-master replication setup for solr (using the java replication in 1.4). My goal here is providing failover for updates, not load balancing. Only one node would receive updates at any given time.
My first attempt didn't work. With a little experimenting and looking around, I saw that in SnapPuller.java, it only checks whether the index version on slave is equal to the version on the master. Even if the slave has a newer index version, it pulls the master's (older) version. It's straightforward to add a few lines to check whether the slave has a later version than the master and then not pull in that case. Obviously, this does not even attempt to deal with load balanced or distributed configuration where any node could get updates. That's not what I'm trying to do. What do the people on the list think about this? (How big of a mistake am I making? / What am I missing?) thanks, Tim --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
