Hi , I am execution pyspark on yarn. I have successfully executed initial dataset but now I growed it 10 times more.
during execution I got all the time this error: 14/09/17 19:28:50 ERROR cluster.YarnClientClusterScheduler: Lost executor 68 on UCS-NODE1.sms1.local: remote Akka client disassociated tasks are failed a resubmitted again: 14/09/17 18:40:42 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 21, 23, 26, 29, 32, 33, 48, 75, 86, 91, 93, 94 14/09/17 18:44:18 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 31, 52, 60, 93 14/09/17 18:46:33 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 19, 20, 23, 27, 39, 51, 64 14/09/17 18:48:27 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 51, 68, 80 14/09/17 18:50:47 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 1, 20, 34, 42, 61, 67, 77, 81, 91 14/09/17 18:58:50 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 8, 21, 23, 29, 34, 40, 46, 67, 69, 86 14/09/17 19:00:44 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 6, 13, 15, 17, 18, 19, 23, 32, 38, 39, 44, 49, 53, 54, 55, 56, 57, 59, 68, 74, 81, 85, 89 14/09/17 19:06:24 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 20, 43, 59, 79, 92 14/09/17 19:16:13 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 0, 2, 3, 11, 24, 31, 43, 65, 73 14/09/17 19:27:40 INFO scheduler.DAGScheduler: Resubmitting Stage 1 (RDD at PythonRDD.scala:252) because some of its tasks had failed: 3, 7, 41, 72, 75, 84 *QUESTION:* how to debug / tune the problem. What can cause to such behavior? I have 5 machine cluster with 32 GB ram. Dataset - 3G. command for execution: /usr/lib/spark-1.0.1.2.1.3.0-563-bin-2.4.0.2.1.3.0-563/bin/spark-submit --master yarn --num-executors 12 --driver-memory 4g --executor-memory 2g --py-files tad.zip --executor-cores 4 /usr/lib/cad/PrepareDataSetYarn.py /input/tad/inpuut.csv /output/cad_model_500_2 Where can I find description of the parameters? --num-executors 12 --driver-memory 4g --executor-memory 2g What parameters should be used for tuning? Thanks Oleg.