[
https://issues.apache.org/jira/browse/HBASE-15146?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15110975#comment-15110975
]
Elliott Clark commented on HBASE-15146:
---------------------------------------
I have some thread dumps (I'll get them when I get to work). They are a little
deceptive since they are only blocking 5 threads. However since those are the
threads feeding everything it's like everything is slow.
bq.If queue is 'too big', we go back to the client. Client retries?
Yeah client retries.
bq.You have it running over there? How does it look when we hit the limit?
Yeah have it running with the first version. However a co-worker found the bug
about not reducing queue size. So need to get that up.
> Don't block on Reader threads queueing to a scheduler queue
> -----------------------------------------------------------
>
> Key: HBASE-15146
> URL: https://issues.apache.org/jira/browse/HBASE-15146
> Project: HBase
> Issue Type: Bug
> Affects Versions: 1.2.0
> Reporter: Elliott Clark
> Assignee: Elliott Clark
> Priority: Blocker
> Fix For: 1.2.0
>
> Attachments: HBASE-15146.0.patch, HBASE-15146.1.patch,
> HBASE-15146.2.patch, HBASE-15146.3.patch
>
>
> Blocking on the epoll thread is awful. The new rpc scheduler can have lots of
> different queues. Those queues have different capacity limits. Currently the
> dispatch method can block trying to add the the blocking queue in any of the
> schedulers.
> This causes readers to block, tcp acks are delayed, and everything slows down.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
