[
https://issues.apache.org/jira/browse/HBASE-7336?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13529682#comment-13529682
]
Lars Hofhansl commented on HBASE-7336:
--------------------------------------
tryLock without a timeout returns immediately. I think in this case it should,
waiting here would potentially wait (at least a threads context switch) on each
single next or get request.
I think this issue would be most prominent in long scans, where the scans
typically do not want to mess up the block cache.
Multiple readers is probably the right solution. How many, though? And I am not
sure how this would interact with the block cache (would probably be OK).
> HFileBlock.readAtOffset does not work well with multiple threads
> ----------------------------------------------------------------
>
> Key: HBASE-7336
> URL: https://issues.apache.org/jira/browse/HBASE-7336
> Project: HBase
> Issue Type: Bug
> Reporter: Lars Hofhansl
> Assignee: Lars Hofhansl
> Priority: Critical
> Fix For: 0.96.0, 0.94.4
>
> Attachments: 7336-0.94.txt, 7336-0.96.txt
>
>
> HBase grinds to a halt when many threads scan along the same set of blocks
> and neither read short circuit is nor block caching is enabled for the dfs
> client ... disabling the block cache makes sense on very large scans.
> It turns out that synchronizing in istream in HFileBlock.readAtOffset is the
> culprit.
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators
For more information on JIRA, see: http://www.atlassian.com/software/jira
