[
https://issues.apache.org/jira/browse/HBASE-22387?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16849309#comment-16849309
]
Anoop Sam John commented on HBASE-22387:
----------------------------------------
bq.I caught a jstack (see blocksByHFile-stack-trace.png) which said
blocksByHFile.add(cacheKey) in BucketCache#cacheBlockWithWaitInternal seems to
be no so efficient when in high Get throughtput.
In fact this blocksByHFile data structure is causing us more heap memory also
when the BC size is much large (Like > 100 GB). In some other issue I was
trying to even get rid of this. We can discuss the pros and cons of that.
> Evaluate the get/scan performance after reading HFile block into offheap
> directly
> ---------------------------------------------------------------------------------
>
> Key: HBASE-22387
> URL: https://issues.apache.org/jira/browse/HBASE-22387
> Project: HBase
> Issue Type: Sub-task
> Reporter: Zheng Hu
> Assignee: Zheng Hu
> Priority: Major
> Attachments: blockReaderRemote.png, blocksByHFile-stack-trace.png,
> screenshot-1.png
>
>
> Now, all sub-tasks has been resolved now (except the HBASE-21946 because of
> the hadoop dependency problem), will provide some performance benckmarks to
> show the latency improvement.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
