[
https://issues.apache.org/jira/browse/HADOOP-16912?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17059050#comment-17059050
]
Fengnan Li commented on HADOOP-16912:
-------------------------------------
Thanks for the reivew [~csun], I addressed all comments except for
{quote}2. in MetricsRegistry#tag you passed ns as port, but it seems it is more
than that. Maybe we should use something different.
{quote}
the ns is basically as `ipc.8020` in reality so that's why I tagged it as port.
> Emit per priority rpc queue time and processing time from DecayRpcScheduler
> ---------------------------------------------------------------------------
>
> Key: HADOOP-16912
> URL: https://issues.apache.org/jira/browse/HADOOP-16912
> Project: Hadoop Common
> Issue Type: New Feature
> Components: common
> Reporter: Fengnan Li
> Assignee: Fengnan Li
> Priority: Major
> Labels: metrics
> Attachments: HADOOP-16912.001.patch, HADOOP-16912.002.patch,
> HADOOP-16912.003.patch
>
>
> At ipc Server level we have the overall rpc queue time and processing time
> for the whole CallQueueManager. In the case of using FairCallQueue, it will
> be great to know the per queue/priority level rpc queue time since many times
> we want to keep certain queues to meet some queue time SLA for customers.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
