[
https://issues.apache.org/jira/browse/HBASE-15525?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15218459#comment-15218459
]
Anoop Sam John commented on HBASE-15525:
----------------------------------------
Ya we overflow on to heap area.. Ya I need to add some logs in the new pool to
tell the effectiveness of the pool.
Yes possibly there can be more than more than one user of the pool and as the
pool BBs are fixed sized one instance is ok.
ya we wanted to read the reqs into pooled BBs (Reuse request read BB, another
jira). Experimenting this but as said, the GC is bit tricky here.
I posted the WIP patch for initial high level review. Let me fix the known
probs and the one [~dvdreddy] pointed out. Ya am expecting help from him to
test this in his cluster as well.. :-)
> OutOfMemory could occur when using BoundedByteBufferPool during RPC bursts
> --------------------------------------------------------------------------
>
> Key: HBASE-15525
> URL: https://issues.apache.org/jira/browse/HBASE-15525
> Project: HBase
> Issue Type: Sub-task
> Components: IPC/RPC
> Reporter: deepankar
> Assignee: Anoop Sam John
> Priority: Critical
> Attachments: WIP.patch
>
>
> After HBASE-13819 the system some times run out of direct memory whenever
> there is some network congestion or some client side issues.
> This was because of pending RPCs in the RPCServer$Connection.responseQueue
> and since all the responses in this queue hold a buffer for cellblock from
> BoundedByteBufferPool this could takeup a lot of memory if the
> BoundedByteBufferPool's moving average settles down towards a higher value
> See the discussion here
> [HBASE-13819-comment|https://issues.apache.org/jira/browse/HBASE-13819?focusedCommentId=15207822&page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel#comment-15207822]
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
