[ https://issues.apache.org/jira/browse/HBASE-19320?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16292059#comment-16292059 ]
Anoop Sam John commented on HBASE-19320: ---------------------------------------- I see.. This is per BB size. hmm.... bq. So the max possible DBB size which can be cached in single ThreadLocal is maxCacheSize * maxCachedBufferSize (worst case) And the max possible total DBB size that the NIO can pool/cache can be calculated as maxCacheSize * maxCachedBufferSize * max possible ThreadLocals max possible ThreadLocals = Reader threads in RpcServer. This can be configured using 'hbase.ipc.server.read.threadpool.size'. Default is 10 (Checked 2.0 only not sure 1.x having a diff default) Just making myself learn this full context and adding that same here for ref also :-) Tks for the sharing [~huaxiang]. > document the mysterious direct memory leak in hbase > ---------------------------------------------------- > > Key: HBASE-19320 > URL: https://issues.apache.org/jira/browse/HBASE-19320 > Project: HBase > Issue Type: Improvement > Affects Versions: 2.0.0, 1.2.6 > Reporter: huaxiang sun > Assignee: huaxiang sun > Attachments: HBASE-19320-master-v001.patch, Screen Shot 2017-11-21 at > 4.43.36 PM.png, Screen Shot 2017-11-21 at 4.44.22 PM.png > > > Recently we run into a direct memory leak case, which takes some time to > trace and debug. Internally discussed with our [~saint....@gmail.com], we > thought we had some findings and want to share with the community. > Basically, it is the issue described in > http://www.evanjones.ca/java-bytebuffer-leak.html and it happened to one of > our hbase clusters. > Create the jira first and will fill in more details later. -- This message was sent by Atlassian JIRA (v6.4.14#64029)