[ https://issues.apache.org/jira/browse/MAPREDUCE-3671?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13185993#comment-13185993 ]
Vinod Kumar Vavilapalli commented on MAPREDUCE-3671: ---------------------------------------------------- .. And this has performance implications on benchmarks. Hurts execution time real bad when there is no speculation for jobs with small tasks. Which is the default. > AM-NM RPC calls occasionally takes a long time to respond > --------------------------------------------------------- > > Key: MAPREDUCE-3671 > URL: https://issues.apache.org/jira/browse/MAPREDUCE-3671 > Project: Hadoop Map/Reduce > Issue Type: Bug > Components: mrv2, nodemanager > Affects Versions: 0.23.0 > Reporter: Siddharth Seth > > Observed while looking at MAPREDUCE-3596 and MAPREDUCE-3656. > startContainer taking over a minute in some cases, otherwise 15 seconds. Both > were observed soon after reduce tasks started. Network congestion ? Need more > looking into. -- 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