[
https://issues.apache.org/jira/browse/YARN-2962?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=15200170#comment-15200170
]
Varun Saxena commented on YARN-2962:
------------------------------------
bq. Could you also explain in the parameter description why one would want to
change it from the default of 0 and how to know what a good split value would
be?
Ok...
bq. I'm not sure this constant adds anything. I found it made the code harder
to read than just hard-coding in 0.
Hmm...If its harder to read, I can put 0 everywhere. This value should not
change in future.
bq. This violates the Principle of Least Astonishment. At least log a warning
that you're not doing what the user said to.
Correct, a warning log should be added.
bq. I don't think the accessors are needed.
Yes, they are not required.
bq. Might want to swap those method names.
Agree. safeDeleteIfExists makes more sense in the other case.
bq. should be HashMap<ApplicationAttemptId, RMAppAttempt> attempts = new
HashMap<>();
Yeah, <> can be used to reduce clutter.
Regarding other comments, will add more comments to make tests and main code
more easier to read and fix missing javadocs.
> ZKRMStateStore: Limit the number of znodes under a znode
> --------------------------------------------------------
>
> Key: YARN-2962
> URL: https://issues.apache.org/jira/browse/YARN-2962
> Project: Hadoop YARN
> Issue Type: Improvement
> Components: resourcemanager
> Affects Versions: 2.6.0
> Reporter: Karthik Kambatla
> Assignee: Varun Saxena
> Priority: Critical
> Attachments: YARN-2962.01.patch, YARN-2962.04.patch,
> YARN-2962.2.patch, YARN-2962.3.patch
>
>
> We ran into this issue where we were hitting the default ZK server message
> size configs, primarily because the message had too many znodes even though
> they individually they were all small.
--
This message was sent by Atlassian JIRA
(v6.3.4#6332)
