hadoop-common-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Amar Kamat (JIRA)" <j...@apache.org>
Subject [jira] Assigned: (HADOOP-3712) TaskTracker can skip a dfs check on every task launch.
Date Tue, 08 Jul 2008 07:49:39 GMT

     [ https://issues.apache.org/jira/browse/HADOOP-3712?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]

Amar Kamat reassigned HADOOP-3712:
----------------------------------

    Assignee: Amar Kamat

> TaskTracker can skip a dfs check on every task launch.
> ------------------------------------------------------
>
>                 Key: HADOOP-3712
>                 URL: https://issues.apache.org/jira/browse/HADOOP-3712
>             Project: Hadoop Core
>          Issue Type: Improvement
>          Components: mapred
>            Reporter: Amar Kamat
>            Assignee: Amar Kamat
>
> When tasktracker gets a new task to run, it queries the namenode to find out the job
directory _size_. This _size_ is required if the job files are not yet localized. But if its
already localized then the dfs call gets wasted. This will be true for all but 1 tasks that
run for a job on a tracker. One can avoid it but checking if the job is already localized.

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


Mime
View raw message