Mailing-List: contact hdfs-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: hdfs-issues@hadoop.apache.org
Date: Fri, 25 Jul 2014 01:03:15 +0000 (UTC)
From: "Allen Wittenauer (JIRA)" <jira@apache.org>
To: hdfs-issues@hadoop.apache.org
Message-ID: <JIRA.12729367.1406182656048.41963.1406250195454@arcas>
In-Reply-To: <JIRA.12729367.1406182656048@arcas>
References: <JIRA.12729367.1406182656048@arcas>
Subject: [jira] [Commented] (HDFS-6747) Display the most recent GC info on
 NN webUI
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HDFS-6747?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=14073916#comment-14073916 ] 

Allen Wittenauer commented on HDFS-6747:
----------------------------------------

If people are using the Web UI to solve those problems, that's bad because it means that we have an education problem.  Most of the pro Hadoop admins (including myself) configure the system such that problems such as these are handled proactively via monitoring and metrics collection.  Putting it on the NN UI is really just a "yup, my phone went off for a valid reason" kind of thing.

Also, as [~esteban] points out, the *length* of the GC is more important that it *happened*. In modern JREs configured with modern GC engines, they do collection all the time.  

 
> Display the most recent GC info on NN webUI
> -------------------------------------------
>
>                 Key: HDFS-6747
>                 URL: https://issues.apache.org/jira/browse/HDFS-6747
>             Project: Hadoop HDFS
>          Issue Type: Improvement
>            Reporter: Ming Ma
>
> It will be handy if the recent GC information is available on NN webUI. admins don't need to dig out GC logs.


--
This message was sent by Atlassian JIRA
(v6.2#6252)