Mailing-List: contact yarn-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Date: Thu, 3 Nov 2016 05:18:58 +0000 (UTC)
From: "Karthik Kambatla (JIRA)" <jira@apache.org>
To: yarn-issues@hadoop.apache.org
Message-ID: <JIRA.13015513.1477504840000.168748.1478150338527@Atlassian.JIRA>
In-Reply-To: <JIRA.13015513.1477504840000@Atlassian.JIRA>
References: <JIRA.13015513.1477504840000@Atlassian.JIRA> <JIRA.13015513.1477504840780@arcas>
Subject: [jira] [Commented] (YARN-5783) Verify applications are identified
 starved
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
archived-at: Thu, 03 Nov 2016 05:19:05 -0000


    [ https://issues.apache.org/jira/browse/YARN-5783?page=3Dcom.atlassian.=
jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=3D15631=
629#comment-15631629 ]=20

Karthik Kambatla commented on YARN-5783:
----------------------------------------

Good points, [~templedf].

Your last suggestion got me thinking. It seemed like mocking the preemption=
 thread to consume starved apps but not actually preempt would allow us to =
test starvation logic better. Did the same. We could potentially add more a=
dvanced tests to verify starvation based on thresholds and timeouts in a su=
bsequent JIRA (YARN-5824).

By the way, this patch conflicts with YARN-5821 which is mostly cosmetic ch=
anges and I believe is ready to go.=20

> Verify applications are identified starved
> ------------------------------------------
>
>                 Key: YARN-5783
>                 URL: https://issues.apache.org/jira/browse/YARN-5783
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: fairscheduler
>    Affects Versions: 2.8.0
>            Reporter: Karthik Kambatla
>            Assignee: Karthik Kambatla
>              Labels: oct16-medium
>         Attachments: yarn-5783.YARN-4752.1.patch, yarn-5783.YARN-4752.2.p=
atch, yarn-5783.YARN-4752.3.patch, yarn-5783.YARN-4752.4.patch, yarn-5783.Y=
ARN-4752.5.patch
>
>
> JIRA to track unit tests to verify the identification of starved applicat=
ions. An application should be marked starved only when:
> # Cluster allocation is over the configured threshold for preemption.
> # Preemption is enabled for a queue and any of the following:
> ## The queue is under its minshare for longer than minsharePreemptionTime=
out
> ## One of the queue=E2=80=99s applications is under its fairshare for lon=
ger than fairsharePreemptionTimeout.


--
This message was sent by Atlassian JIRA
(v6.3.4#6332)

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