Cool,Thanks Mark! Mark Miller-3 wrote > > Either setup a load balancer, or use the SolrCloud solrj client > CloudSolrServer - it takes a comma separated list of zk servers rather > than a solr url. > > On Tue, Sep 11, 2012 at 10:17 PM, andy <yhlweb@> wrote: >> I know fail over is available in solr4.0 right now, if one server >> crashes,other servers also support query,I set up a solr cloud like this >> http://lucene.472066.n3.nabble.com/file/n4007117/Selection_028.png >> >> I use http://localhost:8983/solr/collection1/select?q=*%3A*&wt=xml for >> query >> at first, if the node 8983 crashes, I have to access other nodes for >> query >> like http://localhost:8900/solr/collection1/select?q=*%3A*&wt=xml >> >> but I use the nodes url in the solrj, how to change the request url >> dynamically? >> does SolrCloud support something like virtual ip address? for example I >> use >> url http://collections1 in the solrj, and forward the request to >> available >> url automatically. >> >> >> >> >> -- >> View this message in context: >> http://lucene.472066.n3.nabble.com/SolrCloud-fail-over-tp4007117.html >> Sent from the Solr - User mailing list archive at Nabble.com. > > > > -- > - Mark >
-- View this message in context: http://lucene.472066.n3.nabble.com/SolrCloud-fail-over-tp4007117p4007360.html Sent from the Solr - User mailing list archive at Nabble.com.
