Mailing-List: contact chukwa-dev-help@hadoop.apache.org; run by ezmlm
Precedence: bulk
Reply-To: chukwa-dev@hadoop.apache.org
Message-ID: <189833312.1242841185622.JavaMail.jira@brutus>
Date: Wed, 20 May 2009 10:39:45 -0700 (PDT)
From: "Eric Yang (JIRA)" <jira@apache.org>
To: chukwa-dev@hadoop.apache.org
Subject: [jira] Created: (CHUKWA-256) Change ChukwaJobInstrumentation class
 to use the new API for getting job history file name.
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: 7bit

Change ChukwaJobInstrumentation class to use the new API for getting job history file name.
-------------------------------------------------------------------------------------------

                 Key: CHUKWA-256
                 URL: https://issues.apache.org/jira/browse/CHUKWA-256
             Project: Hadoop Chukwa
          Issue Type: Improvement
          Components: data collection
         Environment: Redhat EL 5.1, Java 6
            Reporter: Eric Yang


Hadoop getJobHistoryFileName has scability problem, the entire method is in a synchronized block.  When there are too many files existed in the log directory, this method takes a very long time to return the file name.  This could cause the job tracker to run very slowly.

There is a new API put in place (HADOOP-4372) getNewJobHistoryFileName should be much more lightweight and work for Chukwa Jobtracker Instrumentation API.

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.