[
https://issues.apache.org/jira/browse/HADOOP-16097?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16765126#comment-16765126
]
Erik Krogen commented on HADOOP-16097:
--------------------------------------
Thanks for the review [~linyiqun]!
I don't quite understand this sentence suggestion:
{quote}
The priority level computation for the users is started from low priority
levels since they will be most common.
{quote}
Can you explain? Why does the user need to be aware of this?
Re: "a RpcMultiplexer" vs. "an RpcMultiplexer," "an" is the correct form here
because it is pronounced using a vowel sound ("arr-pee-see"). See this
[explanation|https://blog.apastyle.org/apastyle/2012/04/using-a-or-an-with-acronyms-and-abbreviations.html].
Re: Line 86, "be proceed" is not correct. Proceed is a verb by itself, there is
no need for "be."
Thanks for reminding me about identity provider, I completely forgot that it is
configurable.
Uploading patch v001, this time with the image.
> Provide proper documentation for FairCallQueue
> ----------------------------------------------
>
> Key: HADOOP-16097
> URL: https://issues.apache.org/jira/browse/HADOOP-16097
> Project: Hadoop Common
> Issue Type: Improvement
> Components: documentation, ipc
> Reporter: Erik Krogen
> Assignee: Erik Krogen
> Priority: Major
> Attachments: HADOOP-16097.000.patch, HADOOP-16097.001.patch,
> faircallqueue-overview.png
>
>
> FairCallQueue, added in HADOOP-10282, doesn't seem to be well-documented
> anywhere. Let's add in a new documentation for it and related components.
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
