[
https://issues.apache.org/jira/browse/HBASE-14331?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14742911#comment-14742911
]
Lars Hofhansl commented on HBASE-14331:
---------------------------------------
Thanks for looking into this [~ikeda]. I usually test these things on just a
single machine (I set up a single node HDFS, and a single node HBase on top of
that). Especially for region server optimization such as this one, that would
be sufficient.
Personally I would prefer a series of smaller patches. I.e. can't we just
optionally replace the LinkedBlockingQueues used with with the SemaphoreBased*
based ones, and then add the nice factories in a 2nd patch? But I see
[~ram_krish] seems to like the other code you added. That's good enough :)
Questions:
Do you think that with this patch we can get rid of the multi call queue logic?
I.e. go back have a single queue (or one for reader and one for writers).
I also really like your suggestion about not waiting when we reached the queue
limit, but return a failure code to the client. We can easily do that with your
semaphore approach.
> a single callQueue related improvements
> ---------------------------------------
>
> Key: HBASE-14331
> URL: https://issues.apache.org/jira/browse/HBASE-14331
> Project: HBase
> Issue Type: Improvement
> Components: IPC/RPC, Performance
> Reporter: Hiroshi Ikeda
> Assignee: Hiroshi Ikeda
> Priority: Minor
> Attachments: BlockingQueuesPerformanceTestApp-output.pdf,
> BlockingQueuesPerformanceTestApp-output.txt,
> BlockingQueuesPerformanceTestApp.java, CallQueuePerformanceTestApp.java,
> HBASE-14331-V2.patch, HBASE-14331-V3.patch, HBASE-14331.patch,
> HBASE-14331.patch, SemaphoreBasedBlockingQueue.java,
> SemaphoreBasedLinkedBlockingQueue.java,
> SemaphoreBasedPriorityBlockingQueue.java
>
>
> {{LinkedBlockingQueue}} well separates locks between the {{take}} method and
> the {{put}} method, but not between takers, and not between putters. These
> methods are implemented to take locks at the almost beginning of their logic.
> HBASE-11355 introduces multiple call-queues to reduce such possible
> congestion, but I doubt that it is required to stick to {{BlockingQueue}}.
> There are the other shortcomings of using {{BlockingQueue}}. When using
> multiple queues, since {{BlockingQueue}} blocks threads it is required to
> prepare enough threads for each queue. It is possible that there is a queue
> starving for threads while there is another queue where threads are idle.
> Even if you can tune parameters to avoid such situations, the tuning is not
> so trivial.
> I suggest using a single {{ConcurrentLinkedQueue}} with {{Semaphore}}.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
