Did you take a look at the node manager logs? Thanks, Ron
Sent from my iPhone > On Mar 23, 2016, at 7:18 AM, Youzha <[email protected]> wrote: > > hi i try to submit mapreduce job but sometimes the nodemanagers are dead when > the job running. i check the hdfs storage only use 20%. and i think the yarn > resource is still enough to run that job. anybody can help me why this can > happen? what requirement that nodemanagers needed? thx before
