[ https://issues.apache.org/jira/browse/SOLR-6312?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14093983#comment-14093983 ]
Steve Davids commented on SOLR-6312: ------------------------------------ Yes, I understand that all updates will always go to the leader, the CPU intensive task in this entire process is running extraction logic using XPaths in the update processor chain before any requests are distributed to the leader/replicas. When the request is distributed to the leader, the leader doesn't need to start the update processor from scratch, instead it continues where the other machine left off in the processing pipeline at the DistributedUpdateProcessor. So if I am able to load balance requests to all replicas the CPU intensive tasks (early update processors) will be shared by multiple machines not just the leader and should result in increased throughput. > CloudSolrServer doesn't honor updatesToLeaders constructor argument > ------------------------------------------------------------------- > > Key: SOLR-6312 > URL: https://issues.apache.org/jira/browse/SOLR-6312 > Project: Solr > Issue Type: Bug > Affects Versions: 4.9 > Reporter: Steve Davids > Fix For: 4.10 > > > The CloudSolrServer doesn't use the updatesToLeaders property - all SolrJ > requests are being sent to the shard leaders. -- This message was sent by Atlassian JIRA (v6.2#6252) --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org