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] Commented: (HADOOP-4934) Distinguish running/successful/failed/killed jobs in jobtracker's history
Date Fri, 30 Jan 2009 13:05:59 GMT

    [ https://issues.apache.org/jira/browse/HADOOP-4934?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=12668855#action_12668855

Amar Kamat commented on HADOOP-4934:

I propose :
# Allow users to group/browse jobs based on various parameters like time(month/week/day/hour),
user, jobname. One way to do it is to store job history files under the user folder, grouped
by date instead of storing it in a flat structure. So a user a will have the jobhistory files
as _job-history-folder/user-name/day/job-history-file_. This will help when there are too
many files in the jobhistory folder.
# Running/Pending/Killed/Failed jobs should not be shown as there is no point to show it in
history. One optimization in the above approach would be to have a 3rd level of hierarchy
in terms of job state. So a running job history file will be stored as _job-history-folder/user-name/day/running/job-history-file_
# By default show last 100 jobs and allow users to change this value via the web-ui

> Distinguish running/successful/failed/killed jobs in jobtracker's history
> -------------------------------------------------------------------------
>                 Key: HADOOP-4934
>                 URL: https://issues.apache.org/jira/browse/HADOOP-4934
>             Project: Hadoop Core
>          Issue Type: Improvement
>          Components: mapred
>    Affects Versions: 0.21.0
>            Reporter: Amar Kamat
>             Fix For: 0.21.0
> It would be nice to group jobs in jobtracker's history based on their completion status.

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

View raw message