[ https://issues.apache.org/jira/browse/MAPREDUCE-4366?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13647200#comment-13647200 ]
Sandy Ryza commented on MAPREDUCE-4366: --------------------------------------- Oof this was a grueling one, but I think I've figured out the problem. At the start of a job, waiting tasks is incremented by the number of tasks, but then decremented each time an attempt launches, even if the attempt is speculative. Also, waiting tasks was incremented whenever a task failed, even if it was a speculative one that failed after a job had completed. This could cause a positive number of waiting tasks to stick around after the job completed. Uploading a patch that fully excludes speculative tasks from waiting tasks. > mapred metrics shows negative count of waiting maps and reduces > --------------------------------------------------------------- > > Key: MAPREDUCE-4366 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-4366 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: jobtracker > Affects Versions: 1.0.2 > Reporter: Thomas Graves > Assignee: Thomas Graves > > Negative waiting_maps and waiting_reduces count is observed in the mapred > metrics. MAPREDUCE-1238 partially fixed this but it appears there is still > issues as we are seeing it, but not as bad. -- This message is automatically generated by JIRA. If you think it was sent incorrectly, please contact your JIRA administrators For more information on JIRA, see: http://www.atlassian.com/software/jira