[ https://issues.apache.org/jira/browse/SOLR-4449?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13582029#comment-13582029 ]
Raintung Li commented on SOLR-4449: ----------------------------------- Hi Philip, I have one suggestion that don't open new thread in BackupRequestLBHttpSolrServer that will increase threads double special in the search function. Search receive request thread will wait the response from multiple shards, this thread can submit the second request to shards of overtime. For example: One search request: 3 shards, the first request: Need 1+3*2 = 7 threads to handle , the second request: Bad case(3 shards all need resend again) 10 threads Change to the first request: Need 1+3=4 threads to handle the second request: Bad case 7 threads Look solr use simple random to do LB. https://blog.heroku.com/archives/2013/2/16/routing_performance_update/ This blog attacked the random strategy was terrible. > Enable backup requests for the internal solr load balancer > ---------------------------------------------------------- > > Key: SOLR-4449 > URL: https://issues.apache.org/jira/browse/SOLR-4449 > Project: Solr > Issue Type: New Feature > Components: SolrCloud > Reporter: philip hoy > Priority: Minor > Attachments: SOLR-4449.patch > > > Add the ability to configure the built-in solr load balancer such that it > submits a backup request to the next server in the list if the initial > request takes too long. Employing such an algorithm could improve the latency > of the 9xth percentile albeit at the expense of increasing overall load due > to additional requests. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira --------------------------------------------------------------------- To unsubscribe, e-mail: dev-unsubscr...@lucene.apache.org For additional commands, e-mail: dev-h...@lucene.apache.org