Mailing-List: contact yarn-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Date: Mon, 31 Jul 2017 20:12:00 +0000 (UTC)
From: "Sunil G (JIRA)" <jira@apache.org>
To: yarn-issues@hadoop.apache.org
Message-ID: <JIRA.13085833.1499657170000.55127.1501531920388@Atlassian.JIRA>
In-Reply-To: <JIRA.13085833.1499657170000@Atlassian.JIRA>
References: <JIRA.13085833.1499657170000@Atlassian.JIRA> <JIRA.13085833.1499657170893@jira-lw-us.apache.org>
Subject: [jira] [Updated] (YARN-6788) Improve performance of resource
 profile branch
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
archived-at: Mon, 31 Jul 2017 20:12:07 -0000


     [ https://issues.apache.org/jira/browse/YARN-6788?page=3Dcom.atlassian=
.jira.plugin.system.issuetabpanels:all-tabpanel ]

Sunil G updated YARN-6788:
--------------------------
    Attachment: YARN-6788-YARN-3926.018.patch

Thanks [~templedf] for comments. Updating new patch fixing same.

bq.=C2=A0What I was trying to say is that=C2=A0getResourceTypesArray()=C2=
=A0starts with a call to=C2=A0getResourceTypes()=C2=A0to make sure the arra=
y is initialized before returning it. Why doesn't=C2=A0getResourceNamesArra=
y()=C2=A0also do that?
In {{DominantResourceCalculator()}}, we already invoke {{getResourceTypes}}=
. So i think its fine.

bq.Sorry to be slow, but I still don't see it. What I mean is that knownRes=
ourceNamesArray is initialized like this
Yes. You are correct,

bq.The risk is that if thread A is just exiting the synchronized block, and=
 thread B is hitting the first check, thread B may see that lock is initial=
ized, but knownResourceNamesArray may not have been flushed yet.
I think i understood the point. But as mentioned in above statement, thread=
 A is just exiting the synchronized block, then resourceTypesArray/NamesArr=
ay must be initialized since its inside synchronized block. Yes, ideally we=
 can optimize here and change to a volatile boolean + array object. If its =
fine, i would like to take that in another ticket so that i can run some th=
read analyzer tools as well. Thoughts?

> Improve performance of resource profile branch
> ----------------------------------------------
>
>                 Key: YARN-6788
>                 URL: https://issues.apache.org/jira/browse/YARN-6788
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: nodemanager, resourcemanager
>            Reporter: Sunil G
>            Assignee: Sunil G
>            Priority: Blocker
>         Attachments: YARN-6788-YARN-3926.001.patch, YARN-6788-YARN-3926.0=
02.patch, YARN-6788-YARN-3926.003.patch, YARN-6788-YARN-3926.004.patch, YAR=
N-6788-YARN-3926.005.patch, YARN-6788-YARN-3926.006.patch, YARN-6788-YARN-3=
926.007.patch, YARN-6788-YARN-3926.008.patch, YARN-6788-YARN-3926.009.patch=
, YARN-6788-YARN-3926.010.patch, YARN-6788-YARN-3926.011.patch, YARN-6788-Y=
ARN-3926.012.patch, YARN-6788-YARN-3926.013.patch, YARN-6788-YARN-3926.014.=
patch, YARN-6788-YARN-3926.015.patch, YARN-6788-YARN-3926.016.patch, YARN-6=
788-YARN-3926.017.patch, YARN-6788-YARN-3926.018.patch
>
>
> Currently we could see a 15% performance delta with this branch.=20
> Few performance improvements to improve the same.
> Also this patch will handle [comments|https://issues.apache.org/jira/brow=
se/YARN-6761?focusedCommentId=3D16075418&page=3Dcom.atlassian.jira.plugin.s=
ystem.issuetabpanels:comment-tabpanel#comment-16075418] from [~leftnoteasy]=
.


--
This message was sent by Atlassian JIRA
(v6.4.14#64029)

---------------------------------------------------------------------
To unsubscribe, e-mail: yarn-issues-unsubscribe@hadoop.apache.org
For additional commands, e-mail: yarn-issues-help@hadoop.apache.org