hadoop-yarn-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Miklos Szegedi (JIRA)" <j...@apache.org>
Subject [jira] [Commented] (YARN-6670) Add separate NM overallocation thresholds for cpu and memory
Date Mon, 12 Jun 2017 17:31:02 GMT

    [ https://issues.apache.org/jira/browse/YARN-6670?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16046813#comment-16046813
] 

Miklos Szegedi commented on YARN-6670:
--------------------------------------

It applies to both CPU and memory. Let's say we have a guaranteed workload that runs in batches.
After each batch it releases some memory, waits for the network for 0.5s and then continues
with high memory and CPU usage. Without a period we will allocate opportunistic containers,
when this happens and preempt them shortly after that. Instead of this, if the CPU and memory
is below the threshold for let's say more then 2 seconds, we would start allocating opportunistic
containers.

> Add separate NM overallocation thresholds for cpu and memory
> ------------------------------------------------------------
>
>                 Key: YARN-6670
>                 URL: https://issues.apache.org/jira/browse/YARN-6670
>             Project: Hadoop YARN
>          Issue Type: Sub-task
>          Components: nodemanager
>    Affects Versions: 3.0.0-alpha3
>            Reporter: Haibo Chen
>            Assignee: Haibo Chen
>         Attachments: YARN-6670-YARN-1011.00.patch, YARN-6670-YARN-1011.01.patch, YARN-6670-YARN-1011.02.patch,
YARN-6670-YARN-1011.03.patch, YARN-6670-YARN-1011.04.patch, YARN-6670-YARN-1011.05.patch
>
>




--
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


Mime
View raw message