[
https://issues.apache.org/jira/browse/HADOOP-16395?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16880740#comment-16880740
]
Erik Krogen commented on HADOOP-16395:
--------------------------------------
If we leave the method as a generic {{close()}}, I don't think it's acceptable
for us to continue using it afterwards. Another implementation may reasonably
assume that after {{close()}} is called, the queue will not be used, and then
do some cleanup steps which would cause the {{put()}} and {{take()}} methods to
no longer work properly. It sounds to me like we need to have something like
{{preclose()}}, {{prepareClose()}}, {{initiateShutdown()}}, {{closeMetrics()}},
etc. to indicate that this isn't closing the queue for all further use. Since
you're leaning towards creating a new interface anyway, I think defining a
method that is more semantically correct makes sense here.
> 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, HADOOP-16395.002.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]
