If i want to take specifically for the task number which got failed. is it possible to take heap dump.
"16/12/16 12:25:54 WARN YarnSchedulerBackend$YarnSchedulerEndpoint: Container killed by YARN for exceeding memory limits. 20.0 GB of 19.8 GB physical memory used. Consider boosting spark.yarn.executor.memoryOverhead. 16/12/16 12:25:54 ERROR YarnClusterScheduler: Lost executor 1 on ip-.dev: Container killed by YARN for exceeding memory limits. 20.0 GB of 19.8 GB physical memory used. Consider boosting spark.yarn.executor.memoryOverhead. 16/12/16 12:25:55 WARN TaskSetManager: Lost task 7.0 in stage 1.0 (TID 9, ip.dev): ExecutorLostFailure (executor 1 exited caused by one of the running tasks) Reason: Container killed by YARN for exceeding memory limits. 20.0 GB of 19.8 GB physical memory used. Consider boosting spark.yarn.executor.memoryOverhead. 16/12/16 12:25:55 INFO BlockManagerMasterEndpoint: Trying to remove executor 1 from BlockManagerMaster. 16/12/16 12:25:55 INFO BlockManagerMaster: Removal of executor 1 requested 16/12/16 12:25:55 INFO YarnSchedulerBackend$YarnDriverEndpoint: Asked to remove non-existent executor 1 " thanks, selvam R On Fri, Dec 16, 2016 at 12:30 PM, Selvam Raman <sel...@gmail.com> wrote: > Hi, > > how can i take heap dump in EMR slave node to analyze. > > I have one master and two slave. > > if i enter jps command in Master, i could see sparksubmit with pid. > > But i could not see anything in slave node. > > how can i take heap dump for spark job. > > -- > Selvam Raman > "லஞ்சம் தவிர்த்து நெஞ்சம் நிமிர்த்து" > -- Selvam Raman "லஞ்சம் தவிர்த்து நெஞ்சம் நிமிர்த்து"