Check the Solr logs for Zookeeper disconnects. It could be that as load is increasing Solr is not able to respond to the Zookeeper pings which would bring the nodes offline. If you see Zookeeper disconnects then you can increase the zkClientTimeout set in solr.xml. But be aware that zk disconnects can also be a sign that your servers are overload and/or under resourced. Memory starvation and stop the world GC can often be the cause of zk disconnects.
On Mon, Apr 1, 2013 at 6:18 AM, sling <sling...@gmail.com> wrote: > hi,all. I am new to Solr. > when i query solrcloud4.1 with solrj, the client throws exceptions as > follows. > there are 2 shards in my solrcloud. > each shard is on a server with 4cpu/3G RAM, and jvm has 2G ram. > when the query requests get more and more, the exception occers. > [java] org.apache.solr.client.solrj.SolrServerException: No live > SolrServers available to handle this request > [java] at > > org.apache.solr.client.solrj.impl.LBHttpSolrServer.request(LBHttpSolrServer.java:486) > [java] at > > org.apache.solr.client.solrj.request.QueryRequest.process(QueryRequest.java:90) > [java] at > org.apache.solr.client.solrj.SolrServer.query(SolrServer.java:301) > [java] at > com.netease.index.service.impl.SearcherServiceImpl.search(Unknown Source) > [java] at com.netease.index.util.ConSearcher.run(Unknown Source) > [java] at > > java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:886) > [java] at > > java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:908) > [java] at java.lang.Thread.run(Thread.java:662) > [java] Caused by: org.apache.solr.client.solrj.SolrServerException: > IOException occured when talking to server at: > http://cms.test.com/solr/doc > [java] at > > org.apache.solr.client.solrj.impl.HttpSolrServer.request(HttpSolrServer.java:416) > [java] at > > org.apache.solr.client.solrj.impl.HttpSolrServer.request(HttpSolrServer.java:181) > [java] at > > org.apache.solr.client.solrj.impl.LBHttpSolrServer.request(LBHttpSolrServer.java:439) > [java] ... 7 more > [java] Caused by: org.apache.http.conn.ConnectionPoolTimeoutException: > Timeout waiting for connection from pool > [java] at > > org.apache.http.impl.conn.tsccm.ConnPoolByRoute.getEntryBlocking(ConnPoolByRoute.java:416) > [java] at > > org.apache.http.impl.conn.tsccm.ConnPoolByRoute$1.getPoolEntry(ConnPoolByRoute.java:299) > [java] at > > org.apache.http.impl.conn.tsccm.ThreadSafeClientConnManager$1.getConnection(ThreadSafeClientConnManager.java:242) > [java] at > > org.apache.http.impl.client.DefaultRequestDirector.execute(DefaultRequestDirector.java:455) > [java] at > > org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:906) > [java] at > > org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:805) > [java] at > > org.apache.http.impl.client.AbstractHttpClient.execute(AbstractHttpClient.java:784) > [java] at > > org.apache.solr.client.solrj.impl.HttpSolrServer.request(HttpSolrServer.java:353) > [java] ... 9 more > > ps: lbhttpsolrserver seems to allocate task imbalance...some node get a > much > heavy load, while others may be not. i use nginx so that task could be > more controllable. is this right? > > > please help me out, Thank you in advance. ^_^ > > > > > > > > > -- > View this message in context: > http://lucene.472066.n3.nabble.com/solr4-1-No-live-SolrServers-available-to-handle-this-request-tp4052862.html > Sent from the Solr - User mailing list archive at Nabble.com. > -- Joel Bernstein Professional Services LucidWorks
