Degree of parallelism in Apache Flink

问题

Can I set different degree of parallelism for different part of the task in our program in Flink? For instance, how does Flink interpret the following sample code? The two custom practitioners MyPartitioner1, MyPartitioner2, partition the input data two 4 and 2 partitions.

partitionedData1 = inputData1
  .partitionCustom(new MyPartitioner1(), 1);
env.setParallelism(4);
DataSet<Tuple2<Integer, Integer>> output1 = partitionedData1
  .mapPartition(new calculateFun());

partitionedData2 = inputData2
  .partitionCustom(new MyPartitioner2(), 2);
env.setParallelism(2);
DataSet<Tuple2<Integer, Integer>> output2 = partitionedData2
  .mapPartition(new calculateFun());

I get the following error for this code:

Exception in thread "main" org.apache.flink.runtime.client.JobExecutionException: Job execution failed.
    at org.apache.flink.runtime.jobmanager.JobManager$$anonfun$receiveWithLogMessages$1.applyOrElse(JobManager.scala:314)
    at scala.runtime.AbstractPartialFunction$mcVL$sp.apply$mcVL$sp(AbstractPartialFunction.scala:33)
    at scala.runtime.AbstractPartialFunction$mcVL$sp.apply(AbstractPartialFunction.scala:33)
    at scala.runtime.AbstractPartialFunction$mcVL$sp.apply(AbstractPartialFunction.scala:25)
    at org.apache.flink.runtime.ActorLogMessages$$anon$1.apply(ActorLogMessages.scala:36)
    at org.apache.flink.runtime.ActorLogMessages$$anon$1.apply(ActorLogMessages.scala:29)
    at scala.PartialFunction$class.applyOrElse(PartialFunction.scala:118)
    at org.apache.flink.runtime.ActorLogMessages$$anon$1.applyOrElse(ActorLogMessages.scala:29)
    at akka.actor.Actor$class.aroundReceive(Actor.scala:465)
    at org.apache.flink.runtime.jobmanager.JobManager.aroundReceive(JobManager.scala:92)
    at akka.actor.ActorCell.receiveMessage(ActorCell.scala:516)
    at akka.actor.ActorCell.invoke(ActorCell.scala:487)
    at akka.dispatch.Mailbox.processMailbox(Mailbox.scala:254)
    at akka.dispatch.Mailbox.run(Mailbox.scala:221)
    at akka.dispatch.Mailbox.exec(Mailbox.scala:231)
    at scala.concurrent.forkjoin.ForkJoinTask.doExec(ForkJoinTask.java:260)
    at scala.concurrent.forkjoin.ForkJoinPool$WorkQueue.runTask(ForkJoinPool.java:1339)
    at scala.concurrent.forkjoin.ForkJoinPool.runWorker(ForkJoinPool.java:1979)
    at scala.concurrent.forkjoin.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:107)
Caused by: java.lang.ArrayIndexOutOfBoundsException: 2
    at org.apache.flink.runtime.io.network.api.writer.RecordWriter.emit(RecordWriter.java:80)
    at org.apache.flink.runtime.operators.shipping.OutputCollector.collect(OutputCollector.java:65)
    at org.apache.flink.runtime.operators.NoOpDriver.run(NoOpDriver.java:92)
    at org.apache.flink.runtime.operators.RegularPactTask.run(RegularPactTask.java:496)
    at org.apache.flink.runtime.operators.RegularPactTask.invoke(RegularPactTask.java:362)
    at org.apache.flink.runtime.taskmanager.Task.run(Task.java:559)
    at java.lang.Thread.run(Unknown Source)

回答1:

ExecutionEnvironment.setParallelism() sets the parallelism for the whole program, i.e., all operators of the program.

You can specify the parallelism for each individual operator by calling the setParallelism() method on the operator.

The ArrayIndexOutOfBoundsException is thrown because your custom partitioner returns an invalid partition number probably due to the unexpected degree of parallelism. The custom partitioner receives the actual parallelism of the receiver as a parameter in its partition(K key, int numPartitions) method.

来源：https://stackoverflow.com/questions/34047548/degree-of-parallelism-in-apache-flink

标签

apache-flink