It's possible that you're seeing a different problem now. I know of at least one bug that can cause the JobTracker to hang for extensive periods of time. It's caused by holding a lock while writing user history to HDFS. MAPREDUCE-5606 tracks this bug, and it actually might be a duplicate of MAPREDUCE-1144.
https://issues.apache.org/jira/browse/MAPREDUCE-5606 https://issues.apache.org/jira/browse/MAPREDUCE-1144 See MAPREDUCE-5606 for a full description of the symptoms. If you run jstack on your JobTracker process, and the thread dump looks similar, then you'll be able to tell if this is the problem. MAPREDUCE-5606 also describes a workaround that you can apply in mapred-site.xml. Chris Nauroth Hortonworks http://hortonworks.com/ On Fri, Nov 15, 2013 at 1:57 PM, Ted Yu <yuzhih...@gmail.com> wrote: > From the command line, can you run 'jmap -heap' ? > > http://download.oracle.com/javase/1.5.0/docs/tooldocs/share/jmap.html > > > On Fri, Nov 15, 2013 at 10:50 AM, Viswanathan J < > jayamviswanat...@gmail.com> wrote: > >> Hi guys, >> >> I had JT OOME in hadoop version 1.2.1 and applied the patch based on the >> fix given by Apache contributors for jira issue mapreduce-5508. >> >> After applying that fix the heap size gradually increases but after one >> week jobtracker completely hangs and slowdown but didn't get JT OOME. No >> error in JT logs. >> >> Without this fix weekly once JT heap size increases and got OOME. >> >> Please help its our production environment. >> >> Thanks, >> > > -- CONFIDENTIALITY NOTICE NOTICE: This message is intended for the use of the individual or entity to which it is addressed and may contain information that is confidential, privileged and exempt from disclosure under applicable law. If the reader of this message is not the intended recipient, you are hereby notified that any printing, copying, dissemination, distribution, disclosure or forwarding of this communication is strictly prohibited. If you have received this communication in error, please contact the sender immediately and delete it from your system. Thank You.