The MRAppmaster process killed count is being added to the Apps Pending value(with -ve sign) in Cluster Metrics page.
---------------------------------------------------------------------------------------------------------------------

                 Key: MAPREDUCE-4122
                 URL: https://issues.apache.org/jira/browse/MAPREDUCE-4122
             Project: Hadoop Map/Reduce
          Issue Type: Bug
          Components: mrv2
    Affects Versions: 0.23.2
            Reporter: Ramgopal N


Application will retry "yarn.resourcemanager.am.max-retries" times before the job is failed,if the MRAppmaster process is getting killed continously.This killed count is considered for Pending applications with -ve value on CLuster metrics page.
  This will mis-interpret the exact number of jobs in the Pending state for the cluster.Even if the MRAppmaster kill count is monitored:should be done at the job level and not at the cluster level.

--
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