Mailing-List: contact common-issues-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: common-issues@hadoop.apache.org
Date: Thu, 21 Jun 2012 20:24:42 +0000 (UTC)
From: "Robert Joseph Evans (JIRA)" <jira@apache.org>
To: common-issues@hadoop.apache.org
Message-ID: <1340162810.40990.1340310282795.JavaMail.jiratomcat@issues-vm>
In-Reply-To: <746372576.39678.1340297743801.JavaMail.jiratomcat@issues-vm>
Subject: [jira] [Commented] (HADOOP-8524) Allow users to get source of a
 Configuration parameter
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit


    [ https://issues.apache.org/jira/browse/HADOOP-8524?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13398822#comment-13398822 ] 

Robert Joseph Evans commented on HADOOP-8524:
---------------------------------------------

The code itself looks good to me.  The only issue I have has nothing to do with your code.  In the past I have talked with others about improving the debuggability/traceability of these values.  For Map/Reduce jobs the config is produced and then written out to job.xml.  After that the config will only look like all of the settings came from job.xml. We lose traceability to the original configurations.  This is even worse for the job.xml that gets sent to the job history server, because it is written out yet again after being read in by the AM, so even the comments in there only say job.xml.

I don't really expect you to fix this in your patch, unless you really want to.  It would just be nice to have the API marked as @Unstable so that if I ever do get around to putting in better traceability we don't have to change this API.  If you are OK with the idea and marking it unstable I will file a separate JIRA for actually adding in the traceability.
                
> Allow users to get source of a Configuration parameter
> ------------------------------------------------------
>
>                 Key: HADOOP-8524
>                 URL: https://issues.apache.org/jira/browse/HADOOP-8524
>             Project: Hadoop Common
>          Issue Type: Improvement
>          Components: conf
>    Affects Versions: 2.0.0-alpha
>            Reporter: Harsh J
>            Assignee: Harsh J
>            Priority: Trivial
>         Attachments: HADOOP-8524.patch, HADOOP-8524.patch
>
>
> When we load the various XMLs via the Configuration class, the source of the XML file (filename) is usually kept in the Configuration class but not exposed programmatically. It is presently exposed as comments such as "Loaded from mapred-site.xml" in the XML dump/serialization but can't be accessed otherwise (Via the Configuration API).
> For debugging/etc. purposes, it may be useful to expose this safely (such as an API for "where did this property come from?" queries for a specific property, via an API.

--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators: https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira