[
https://issues.apache.org/jira/browse/HBASE-7379?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13536666#comment-13536666
]
Ted Yu commented on HBASE-7379:
-------------------------------
Thanks for the review, Elliott.
Looking at BlockingQueue API, I wonder how we handle point #4 (starvation)
above.
Suppose requests r1, r2 and r3 are in the queue where r1 carries data much
larger than those of r2 and r3. When we take r1 off the queue and discover that
its data size exceeds threshold, do we put it at the end of the queue or put r1
at the head of the queue (after tacking out r2) ?
Would we face similar situation where r1 is continuously pushed off ?
> Port '[89-fb] prevent OOM possibility due to per connection responseQueue
> being unbounded' to trunk
> ---------------------------------------------------------------------------------------------------
>
> Key: HBASE-7379
> URL: https://issues.apache.org/jira/browse/HBASE-7379
> Project: HBase
> Issue Type: Sub-task
> Reporter: Ted Yu
> Assignee: Ted Yu
> Fix For: 0.96.0
>
> Attachments: 7379-trunk.txt, 7379-trunk-v2.txt
>
>
> HBASE-6728 ported '[89-fb] prevent OOM possibility due to per connection
> responseQueue being unbounded' to 0.94 branch
> This issue is to port it to trunk
--
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
