[
https://issues.apache.org/jira/browse/HADOOP-16395?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16874270#comment-16874270
]
Erik Krogen commented on HADOOP-16395:
--------------------------------------
Hi [~hfyang20071], thanks for reporting this. There is one thing I am confused
about with your patch. The comment on {{FairCallQueue.MetricsProxy}} says:
{code}
/**
* MetricsProxy is a singleton because we may init multiple
* FairCallQueues, but the metrics system cannot unregister beans cleanly.
*/
{code}
This claims that we can't unregister beans properly. Does the code in place for
{{DecayRpcScheduler.MetricsProxy}} work properly? If so, do we need to update
this comment on {{FairCallQueue.MetricsProxy}}?
Assuming that this is the right approach, I have some more specific comments on
the patch:
* Why are we defining a new {{CloseableQueue}} interface? Can't we just say
that {{FairCallQueue implements Closeable}}?
* You've added a {{close()}} method to {{FairCallQueue}}, but it is not called
anywhere. I think in {{swapQueues}} you need to check if the queue is
{{instanceof Closeable}}, and if so, close it.
> when CallQueueManager swap queues, we should remove metrics about FCQ
> ---------------------------------------------------------------------
>
> Key: HADOOP-16395
> URL: https://issues.apache.org/jira/browse/HADOOP-16395
> Project: Hadoop Common
> Issue Type: Improvement
> Components: common, ipc
> Affects Versions: 3.2.0
> Reporter: yanghuafeng
> Priority: Minor
> Attachments: HADOOP-16395.001.patch
>
>
> when we use "dfsadmin -refreshCallQueue" between FCQ and LBQ, we find that
> the metrics about FCQ still display by the JMX. Normally the metrics should
> disappear like what DecayRpcScheduler did.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
