SolrCloud OOM heap space
Hi All, I am getting constant OOM errors on a SolrCloud instance. (3 shards, 2 solr instance in each shard, each server with 22gb Of Memory, Xmx = 12GB for java ) . Here is a error log: http://pastie.org/private/dcga3kfatvvamslmtvrp0g As of now Iam not indexing any more documents. The total size of index on each server is around 36-40 GB. -Xmx12288m -DSTOP.PORT=8079 -DSTOP.KEY=ABC -XX:NewRatio=1 -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:+CMSParallelRemarkEnabled -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=/var/log -Djetty.port=8983 -DzkHost=ZooKeeperServer001:2181 -jar start.jar If anyone has faced similar issues please let me know. --Shreejay -- View this message in context: http://lucene.472066.n3.nabble.com/SolrCloud-OOM-heap-space-tp4025821.html Sent from the Solr - User mailing list archive at Nabble.com.
RE: SolrCloud OOM heap space
Hi - the stack trace and preceding log entries look similar to what i've seen and reported on. A patch has just been attached to the issue, perhaps you can try it if the description matches your scenario and report back on Jira. https://issues.apache.org/jira/browse/SOLR-4144 -Original message- From:shreejay shreej...@gmail.com Sent: Mon 10-Dec-2012 23:22 To: solr-user@lucene.apache.org Subject: SolrCloud OOM heap space Hi All, I am getting constant OOM errors on a SolrCloud instance. (3 shards, 2 solr instance in each shard, each server with 22gb Of Memory, Xmx = 12GB for java ) . Here is a error log: http://pastie.org/private/dcga3kfatvvamslmtvrp0g As of now Iam not indexing any more documents. The total size of index on each server is around 36-40 GB. -Xmx12288m -DSTOP.PORT=8079 -DSTOP.KEY=ABC -XX:NewRatio=1 -XX:+UseParNewGC -XX:+UseConcMarkSweepGC -XX:+CMSParallelRemarkEnabled -XX:+HeapDumpOnOutOfMemoryError -XX:HeapDumpPath=/var/log -Djetty.port=8983 -DzkHost=ZooKeeperServer001:2181 -jar start.jar If anyone has faced similar issues please let me know. --Shreejay -- View this message in context: http://lucene.472066.n3.nabble.com/SolrCloud-OOM-heap-space-tp4025821.html Sent from the Solr - User mailing list archive at Nabble.com.
RE: SolrCloud OOM heap space
Thanks Markus. Is this issue only on 4.x and 5.x branches? I am currently running a v recent build of 4.x branch with an applied patch. I just want to make sure that this is not an issue with 4.0. In which case I can think of applying my patch to 4.0 instead of 4x or 5x. --Shreejay -- View this message in context: http://lucene.472066.n3.nabble.com/SolrCloud-OOM-heap-space-tp4025821p4025839.html Sent from the Solr - User mailing list archive at Nabble.com.
RE: SolrCloud OOM heap space
Hi - We're using trunk (5x) but we don't see it on trunk builds from a few months ago. In the case of the linked issue the oom occurs some time after start up but i'm not sure this applies to you. You can test the patch if you think it applies to you, we will test it tomorrow. If the patch does not help you or you experience a different issue you might want to open a new issue. -Original message- From:shreejay shreej...@gmail.com Sent: Tue 11-Dec-2012 00:31 To: solr-user@lucene.apache.org Subject: RE: SolrCloud OOM heap space Thanks Markus. Is this issue only on 4.x and 5.x branches? I am currently running a v recent build of 4.x branch with an applied patch. I just want to make sure that this is not an issue with 4.0. In which case I can think of applying my patch to 4.0 instead of 4x or 5x. --Shreejay -- View this message in context: http://lucene.472066.n3.nabble.com/SolrCloud-OOM-heap-space-tp4025821p4025839.html Sent from the Solr - User mailing list archive at Nabble.com.
RE: SolrCloud OOM heap space
Thanks Marcus. I will apply the patch to the 4x branch I have, and report back. -- View this message in context: http://lucene.472066.n3.nabble.com/SolrCloud-OOM-heap-space-tp4025821p4025858.html Sent from the Solr - User mailing list archive at Nabble.com.
