Very few GC messages in the logs, and none around the OOM instances... Cheers, -Robin-
On Wednesday, 25 June 2014 16:55:03 UTC+2, Michael Hart wrote: > > What does your GC Old Count and GC Old Duration look like? Do you have > warnings in the logs about long GC's? > I've got similar issues and telltale sign of when things are about to go > south is when the old GC count starts to rise, and GC old duration > increases. > > On Wednesday, June 25, 2014 9:54:27 AM UTC-4, Robin Clarke wrote: >> >> I have a 10 machine cluster where frequently (about once per day when >> indexing and querying is at its height) one elasticsearch node goes OOM... >> It usually recovers, but by this time the cluster is redistributing the >> lost shards, which causes more load, which often in turn causes an OOM on >> another machine. >> Each machine has 32GB memory of which I currently have 12GB allocated to >> Elasticsearch. I have logstash (max 500M) and redis (max 2GB) running on >> the machines too, and see that the remaining ~17GB is used for file >> cache... i.e. it all looks healthy, up until the moment when elasticsearch >> spews e.g. this sequence of errors: >> >> >> Actual Exception >> org.elasticsearch.search.query.QueryPhaseExecutionException: >> [logstash-2014.06.24][1]: query[ConstantScore(*:*)],from[0],size[0]: Query >> Failed [Failed to execute main query] at >> org.elasticsearch.search.query.QueryPhase.execute(QueryPhase.java:127) at >> org.elasticsearch.search.SearchService.executeQueryPhase(SearchService.java:257) >> >> at >> org.elasticsearch.search.action.SearchServiceTransportAction$SearchQueryTransportHandler.messageReceived(SearchServiceTransportAction.java:623) >> >> at >> org.elasticsearch.search.action.SearchServiceTransportAction$SearchQueryTransportHandler.messageReceived(SearchServiceTransportAction.java:612) >> >> at >> org.elasticsearch.transport.netty.MessageChannelHandler$RequestHandler.run(MessageChannelHandler.java:270) >> >> at >> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) >> >> at >> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) >> >> at java.lang.Thread.run(Thread.java:744) Caused by: >> java.lang.OutOfMemoryError: Java heap space >> >> >> Failed to send error message back to client for action >> [search/phase/query] java.lang.OutOfMemoryError: Java heap space >> >> >> Actual Exception org.elasticsearch.index.IndexShardMissingException: >> [logstash-2014.06.25][3] missing at >> org.elasticsearch.index.service.InternalIndexService.shardSafe(InternalIndexService.java:182) >> >> at >> org.elasticsearch.search.SearchService.createContext(SearchService.java:496) >> at >> org.elasticsearch.search.SearchService.createAndPutContext(SearchService.java:480) >> >> at >> org.elasticsearch.search.SearchService.executeQueryPhase(SearchService.java:252) >> >> at >> org.elasticsearch.search.action.SearchServiceTransportAction$SearchQueryTransportHandler.messageReceived(SearchServiceTransportAction.java:623) >> >> at >> org.elasticsearch.search.action.SearchServiceTransportAction$SearchQueryTransportHandler.messageReceived(SearchServiceTransportAction.java:612) >> >> at >> org.elasticsearch.transport.netty.MessageChannelHandler$RequestHandler.run(MessageChannelHandler.java:270) >> >> at >> java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1145) >> >> at >> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:615) >> >> at java.lang.Thread.run(Thread.java:744) >> >> Any ideas what might be going wrong here, or what I might be able to do >> to remedy the situation? >> >> Cheers, >> -Robin- >> > -- You received this message because you are subscribed to the Google Groups "elasticsearch" group. To unsubscribe from this group and stop receiving emails from it, send an email to elasticsearch+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/elasticsearch/66d2dcc6-5ed0-4032-bbf7-b49ca662dbd5%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.