Я выполняю задание spark-scala с использованием spark-shell, и проблема, с которой я сталкиваюсь, заключается в том, что в конце заключительного этапа и окончательного сопоставления, как на этапе 5, он выделяет 50 и завершает 49 очень быстро, а на 50-мзанимает 5 минут и говорит, что не хватает памяти и выходит из строя.Я использую SPARK_MAJOR_VERSION=2
Я использую приведенную ниже команду spark-shell --master yarn --conf spark.driver.memory=30G --conf spark.executor.memory=40G --conf spark.shuffle.service.enabled=true --conf spark.dynamicAllocation.enabled=false --conf spark.sql.broadcastTimeout=36000 --conf spark.shuffle.compress=true --conf spark.executor.heartbeatInterval=3600s --conf spark.executor.instance=160
В приведенной выше конфигурации я попытался динамически выделить значение true и запустил память драйвера и исполнителя из 1 ГБ.У меня общая оперативная память 6,78 ТБ и 1300 VCores (это все мое аппаратное обеспечение hadoop).
Таблица, которую я читаю, 40GB, и я объединяю 6 таблиц с этой таблицей 40 ГБ, так что в целомбыть 60 ГБ.таким образом, искра инициализирует 4 этапа для этого, а на последнем этапе в конце она терпит неудачу.Я использую spark sql для выполнения SQL.
Ниже приведены ошибки:
19/04/26 14:29:02 WARN HeartbeatReceiver: Removing executor 2 with no recent heartbeats: 125967 ms exceeds timeout 120000 ms
19/04/26 14:29:02 ERROR YarnScheduler: Lost executor 2 on worker03.some.com: Executor heartbeat timed out after 125967 ms
19/04/26 14:29:02 WARN TaskSetManager: Lost task 5.0 in stage 2.0 (TID 119, worker03.some.com, executor 2): ExecutorLostFailure (executor 2 exited caused by one of the running tasks) Reason: Executor heartbeat timed out after 125967 ms
19/04/26 14:29:02 WARN HeartbeatReceiver: Removing executor 1 with no recent heartbeats: 126225 ms exceeds timeout 120000 ms
19/04/26 14:29:02 ERROR YarnScheduler: Lost executor 1 on ncednhpwrka0008.devhadoop.charter.com: Executor heartbeat timed out after 126225 ms
19/04/26 14:29:02 WARN YarnSchedulerBackend$YarnSchedulerEndpoint: Container marked as failed: container_e1223_1556277056929_0976_01_000003 on host: worker03.some.com. Exit status: 52. Diagnostics: Exception from container-launch.
Container id: container_e1223_1556277056929_0976_01_000003
Exit code: 52
Shell output: main : command provided 1
main : run as user is svc-bd-xdladmrw-dev
main : requested yarn user is svc-bd-xdladmrw-dev
Getting exit code file...
Creating script paths...
Writing pid file...
Writing to tmp file /data/00/yarn/local/nmPrivate/application_1556277056929_0976/container_e1223_1556277056929_0976_01_000003/container_e1223_1556277056929_0976_01_000003.pid.tmp
Writing to cgroup task files...
Creating local dirs...
Launching container...
Getting exit code file...
Creating script paths...
Container exited with a non-zero exit code 52. Last 4096 bytes of stderr :
0 in stage 2.0 (TID 119)
19/04/26 14:27:37 INFO HadoopRDD: Input split: hdfs://datadev/data/dev/HIVE_SCHEMA/somedb.db/sbscr_usge_cycl_key_xref/000000_0_copy_2:0+6623042
19/04/26 14:27:37 INFO OrcRawRecordMerger: min key = null, max key = null
19/04/26 14:27:37 INFO ReaderImpl: Reading ORC rows from hdfs://datadev/data/dev/HIVE_SCHEMA/somedb.db/sbscr_usge_cycl_key_xref/000000_0_copy_2 with {include: [true, true, true], offset: 0, length: 9223372036854775807}
19/04/26 14:29:00 ERROR Executor: Exception in task 5.0 in stage 2.0 (TID 119)
java.lang.OutOfMemoryError
at java.io.ByteArrayOutputStream.hugeCapacity(ByteArrayOutputStream.java:123)
at java.io.ByteArrayOutputStream.grow(ByteArrayOutputStream.java:117)
at java.io.ByteArrayOutputStream.ensureCapacity(ByteArrayOutputStream.java:93)
at java.io.ByteArrayOutputStream.write(ByteArrayOutputStream.java:153)
at net.jpountz.lz4.LZ4BlockOutputStream.flushBufferedData(LZ4BlockOutputStream.java:205)
at net.jpountz.lz4.LZ4BlockOutputStream.write(LZ4BlockOutputStream.java:158)
at java.io.DataOutputStream.write(DataOutputStream.java:107)
at org.apache.spark.sql.catalyst.expressions.UnsafeRow.writeToStream(UnsafeRow.java:554)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:237)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:228)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:323)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:287)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
at org.apache.spark.scheduler.Task.run(Task.scala:108)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:338)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
19/04/26 14:29:00 ERROR SparkUncaughtExceptionHandler: Uncaught exception in thread Thread[Executor task launch worker for task 119,5,main]
java.lang.OutOfMemoryError
at java.io.ByteArrayOutputStream.hugeCapacity(ByteArrayOutputStream.java:123)
at java.io.ByteArrayOutputStream.grow(ByteArrayOutputStream.java:117)
at java.io.ByteArrayOutputStream.ensureCapacity(ByteArrayOutputStream.java:93)
at java.io.ByteArrayOutputStream.write(ByteArrayOutputStream.java:153)
at net.jpountz.lz4.LZ4BlockOutputStream.flushBufferedData(LZ4BlockOutputStream.java:205)
at net.jpountz.lz4.LZ4BlockOutputStream.write(LZ4BlockOutputStream.java:158)
at java.io.DataOutputStream.write(DataOutputStream.java:107)
at org.apache.spark.sql.catalyst.expressions.UnsafeRow.writeToStream(UnsafeRow.java:554)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:237)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:228)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:323)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:287)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
at org.apache.spark.scheduler.Task.run(Task.scala:108)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:338)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
19/04/26 14:29:00 INFO DiskBlockManager: Shutdown hook called
19/04/26 14:29:00 INFO ShutdownHookManager: Shutdown hook called
19/04/26 14:29:02 ERROR YarnScheduler: Lost executor 2 on worker03.some.com: Container marked as failed: container_e1223_1556277056929_0976_01_000003 on host: worker03.some.com. Exit status: 52. Diagnostics: Exception from container-launch.
Container id: container_e1223_1556277056929_0976_01_000003
Exit code: 52
Shell output: main : command provided 1
main : run as user is svc-bd-xdladmrw-dev
main : requested yarn user is svc-bd-xdladmrw-dev
Getting exit code file...
Creating script paths...
Writing pid file...
Writing to tmp file /data/00/yarn/local/nmPrivate/application_1556277056929_0976/container_e1223_1556277056929_0976_01_000003/container_e1223_1556277056929_0976_01_000003.pid.tmp
Writing to cgroup task files...
Creating local dirs...
Launching container...
Getting exit code file...
Creating script paths...
Container exited with a non-zero exit code 52. Last 4096 bytes of stderr :
0 in stage 2.0 (TID 119)
19/04/26 14:27:37 INFO HadoopRDD: Input split: hdfs://datadev/data/dev/HIVE_SCHEMA/somedb.db/sbscr_usge_cycl_key_xref/000000_0_copy_2:0+6623042
19/04/26 14:27:37 INFO OrcRawRecordMerger: min key = null, max key = null
19/04/26 14:27:37 INFO ReaderImpl: Reading ORC rows from hdfs://datadev/data/dev/HIVE_SCHEMA/somedb.db/sbscr_usge_cycl_key_xref/000000_0_copy_2 with {include: [true, true, true], offset: 0, length: 9223372036854775807}
19/04/26 14:29:00 ERROR Executor: Exception in task 5.0 in stage 2.0 (TID 119)
java.lang.OutOfMemoryError
at java.io.ByteArrayOutputStream.hugeCapacity(ByteArrayOutputStream.java:123)
at java.io.ByteArrayOutputStream.grow(ByteArrayOutputStream.java:117)
at java.io.ByteArrayOutputStream.ensureCapacity(ByteArrayOutputStream.java:93)
at java.io.ByteArrayOutputStream.write(ByteArrayOutputStream.java:153)
at net.jpountz.lz4.LZ4BlockOutputStream.flushBufferedData(LZ4BlockOutputStream.java:205)
at net.jpountz.lz4.LZ4BlockOutputStream.write(LZ4BlockOutputStream.java:158)
at java.io.DataOutputStream.write(DataOutputStream.java:107)
at org.apache.spark.sql.catalyst.expressions.UnsafeRow.writeToStream(UnsafeRow.java:554)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:237)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:228)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:323)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:287)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
at org.apache.spark.scheduler.Task.run(Task.scala:108)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:338)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
19/04/26 14:29:00 ERROR SparkUncaughtExceptionHandler: Uncaught exception in thread Thread[Executor task launch worker for task 119,5,main]
java.lang.OutOfMemoryError
at java.io.ByteArrayOutputStream.hugeCapacity(ByteArrayOutputStream.java:123)
at java.io.ByteArrayOutputStream.grow(ByteArrayOutputStream.java:117)
at java.io.ByteArrayOutputStream.ensureCapacity(ByteArrayOutputStream.java:93)
at java.io.ByteArrayOutputStream.write(ByteArrayOutputStream.java:153)
at net.jpountz.lz4.LZ4BlockOutputStream.flushBufferedData(LZ4BlockOutputStream.java:205)
at net.jpountz.lz4.LZ4BlockOutputStream.write(LZ4BlockOutputStream.java:158)
at java.io.DataOutputStream.write(DataOutputStream.java:107)
at org.apache.spark.sql.catalyst.expressions.UnsafeRow.writeToStream(UnsafeRow.java:554)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:237)
at org.apache.spark.sql.execution.SparkPlan$$anonfun$2.apply(SparkPlan.scala:228)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.RDD$$anonfun$mapPartitionsInternal$1$$anonfun$apply$25.apply(RDD.scala:827)
at org.apache.spark.rdd.MapPartitionsRDD.compute(MapPartitionsRDD.scala:38)
at org.apache.spark.rdd.RDD.computeOrReadCheckpoint(RDD.scala:323)
at org.apache.spark.rdd.RDD.iterator(RDD.scala:287)
at org.apache.spark.scheduler.ResultTask.runTask(ResultTask.scala:87)
at org.apache.spark.scheduler.Task.run(Task.scala:108)
at org.apache.spark.executor.Executor$TaskRunner.run(Executor.scala:338)
at java.util.concurrent.ThreadPoolExecutor.runWorker(ThreadPoolExecutor.java:1149)
at java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:624)
at java.lang.Thread.run(Thread.java:748)
19/04/26 14:29:00 INFO DiskBlockManager: Shutdown hook called
19/04/26 14:29:00 INFO ShutdownHookManager: Shutdown hook called
Может кто-нибудь сообщить мне, если я здесь что-то не так, например, распределение памяти или что-то еще?Пожалуйста, предложите любые альтернативы, чтобы завершить эту работу, не получая исключение нашей памяти или потерянного рабочего узла.Любая помощь или информация с благодарностью.
Спасибо!