Mailing-List: contact mapreduce-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: mapreduce-issues@hadoop.apache.org
Date: Fri, 16 Jan 2015 10:23:35 +0000 (UTC)
From: "Tsuyoshi OZAWA (JIRA)" <jira@apache.org>
To: mapreduce-issues@hadoop.apache.org
Message-ID: <JIRA.12764673.1420245371000.100588.1421403815642@Atlassian.JIRA>
In-Reply-To: <JIRA.12764673.1420245371000@Atlassian.JIRA>
References: <JIRA.12764673.1420245371000@Atlassian.JIRA>
 <JIRA.12764673.1420245371366@arcas>
Subject: [jira] [Commented] (MAPREDUCE-6209) Implement a heuristic to
 auto-size Java heap of MRAppMaster container
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/MAPREDUCE-6209?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14280062#comment-14280062 ] 

Tsuyoshi OZAWA commented on MAPREDUCE-6209:
-------------------------------------------

[~jira.shegalov] Tez is doing similar approach: task memory * 0.8  is used to avoid OoM. It looks good to me. Do you have additional ideas?

> Implement a heuristic to auto-size Java heap of MRAppMaster container
> ---------------------------------------------------------------------
>
>                 Key: MAPREDUCE-6209
>                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-6209
>             Project: Hadoop Map/Reduce
>          Issue Type: Improvement
>          Components: applicationmaster
>            Reporter: Gera Shegalov
>
> The size of Java heap required by the AM is linearly proportional to the size of the MR job (number of mappers/splits and reducers). it would be nice if users did not have to adjust the AM container size when transitioning from testing job on a small sample to a production job on a full-scale dataset. 


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