Mailing-List: contact common-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: common-issues@hadoop.apache.org
Date: Mon, 8 Aug 2011 15:42:27 +0000 (UTC)
From: "Steve Hoffman (JIRA)" <jira@apache.org>
To: common-issues@hadoop.apache.org
Message-ID: 
 <2079531816.16815.1312818147196.JavaMail.tomcat@hel.zones.apache.org>
Subject: [jira] [Commented] (HADOOP-6244) Improvements to FileContext
 metrics output formatting
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HADOOP-6244?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13081007#comment-13081007 ] 

Steve Hoffman commented on HADOOP-6244:
---------------------------------------

I'd like to argue that this should be routed thru log4j so we can use something like RollingFileAppender to deal with the long term consequences of the file being written to forever.
Then the configurable part is just what format you want the record.  Again, can take a hint from log4j by providing %LETTER parameters for specifying that.

The root of the problem I see here is that any simple 'grep' log message by definition has to have 1 line per record with a newline.  Thus, you can't have a structured message in there w/o doing some kind of encoding which leads you to avro/protobufs/etc. which make it not usable in grep (usually).  Seems you can't have it both ways.

This doesn't apply so much to metrics as you can just use key/value pairs on 1 line, but would still like to see this more general so if I WANTED protobufs/avro/etc. I would just implement the record formatter part and output a 'string'

> Improvements to FileContext metrics output formatting
> -----------------------------------------------------
>
>                 Key: HADOOP-6244
>                 URL: https://issues.apache.org/jira/browse/HADOOP-6244
>             Project: Hadoop Common
>          Issue Type: Improvement
>          Components: metrics
>    Affects Versions: 0.22.0
>            Reporter: Todd Lipcon
>            Assignee: Steve Loughran
>         Attachments: hadoop-6244.txt
>
>
> The output of FileContext has two big issues: 1) it doesn't include a timestamp, 2) it doesn't differentiate between tags and metrics in formatting. This patch is to improve the output format to be more useful.

--
This message is automatically generated by JIRA.
For more information on JIRA, see: http://www.atlassian.com/software/jira