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] Created: (HADOOP-5526) Provide an admin page displaying events in the cluster along with cluster status/health
Date Wed, 18 Mar 2009 11:29:51 GMT
Provide an admin page displaying events in the cluster along with cluster status/health
---------------------------------------------------------------------------------------

                 Key: HADOOP-5526
                 URL: https://issues.apache.org/jira/browse/HADOOP-5526
             Project: Hadoop Core
          Issue Type: New Feature
          Components: mapred
            Reporter: Amar Kamat


Here are few things that will help admins understand whats happening in the cluster
# Events updates
  ## recently added tracker
  ## lost trackers
  ## recently submitted jobs
  ## user updates
  ## killed/failed attempts/tasks 
  ## killed jobs and the reason
  ## recent exceptions like oom etc
  ## expired tasks
  ## recovery manager updates
  ## memory/cpu usage
  ## black listing of tracker 
  ## killing of maps based on fetch failures
  ## info about why some jobs was rejected(acls, max tasks)/failed(failures)/killed (user)
  ## etc
# Status :
  ## tracker health and status
  ## User status
    ### num jobs submitted
    ### total time the cluster was used
    ### success/failed/killed history
  ## job status
     ### task completion events
     ### recently scheduled tasks
     ### progress
     ### killed/failed/success history
  ## space on the box where the jt is running
  ## etc
# Config :
  ## slot info
  ## acl info
  ## etc

----
Graphical views and auto updation would be cool. Raising alarms upon certain events would
be super cool.

-- 
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