[
https://issues.apache.org/jira/browse/HBASE-15967?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15317742#comment-15317742
]
stack commented on HBASE-15967:
-------------------------------
Say more [~ikeda] ...
I think we need the workers behind the queue (the 'handler' threads) because we
want to do some scheduling... If priority rpc, it should happen above 'normal'
priority ipc. If we only have a pool of readers, and they do the parse of the
request and then do the handling instead of handing it off, then we don't have
opportunity to 'schedule' (we will run much faster though!) See related
HBASE-15971. Thanks.
> Metric for active ipc Readers and make default fraction of cpu count
> --------------------------------------------------------------------
>
> Key: HBASE-15967
> URL: https://issues.apache.org/jira/browse/HBASE-15967
> Project: HBase
> Issue Type: Sub-task
> Reporter: stack
> Assignee: stack
> Attachments: HBASE-15967.master.001.patch
>
>
> Our ipc Readers are hard coded at 10 regardless since . Running w/ less
> Readers, we go faster..(e.g. 12 Readers has us doing 135k with workloadc and
> 6 readers has us doing 145k).. .but hard to tell what count of Readers are
> needed since no metric.
> This issue changes Readers to be 1/4 the installed CPUs or 8, whichever is
> the minimum, and then adds a new hbase.regionserver.ipc.runningReaders metric
> so you have a chance seeing whats needed.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
