org.apache.beam.sdk.transforms

Class IntraBundleParallelization

java.lang.Object

org.apache.beam.sdk.transforms.IntraBundleParallelization

public class IntraBundleParallelization
extends Object

Provides multi-threading of DoFns, using threaded execution to process multiple elements concurrently within a bundle.

Note, that each Dataflow worker will already process multiple bundles concurrently and usage of this class is meant only for cases where processing elements from within a bundle is limited by blocking calls.

CPU intensive or IO intensive tasks are in general a poor fit for parallelization. This is because a limited resource that is already maximally utilized does not benefit from sub-division of work. The parallelization will increase the amount of time to process each element yet the throughput for processing will remain relatively the same. For example, if the local disk (an IO resource) has a maximum write rate of 10 MiB/s, and processing each element requires to write 20 MiBs to disk, then processing one element to disk will take 2 seconds. Yet processing 3 elements concurrently (each getting an equal share of the maximum write rate) will take at least 6 seconds to complete (there is additional overhead in the extra parallelization).

To parallelize a DoFn to 10 threads:

 PCollection<T> data = ...;
 data.apply(
   IntraBundleParallelization.of(new MyDoFn())
                             .withMaxParallelism(10)));
 

An uncaught exception from the wrapped DoFn will result in the exception being rethrown in later calls to IntraBundleParallelization.MultiThreadedIntraBundleProcessingDoFn.processElement(org.apache.beam.sdk.transforms.DoFn<InputT, OutputT>.ProcessContext) or a call to IntraBundleParallelization.MultiThreadedIntraBundleProcessingDoFn.finishBundle(org.apache.beam.sdk.transforms.DoFn<InputT, OutputT>.Context).

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	IntraBundleParallelization.Bound<InputT,OutputT> A PTransform that, when applied to a PCollection<InputT>, invokes a user-specified DoFn<InputT, OutputT> on all its elements, with all its outputs collected into an output PCollection<OutputT>.
static class	IntraBundleParallelization.MultiThreadedIntraBundleProcessingDoFn<InputT,OutputT> A multi-threaded DoFn wrapper.
static class	IntraBundleParallelization.Unbound An incomplete IntraBundleParallelization transform, with unbound input/output types.

Constructor Summary

Constructors

Constructor and Description
IntraBundleParallelization()

Method Summary

Modifier and Type	Method and Description
static <InputT,OutputT> IntraBundleParallelization.Bound<InputT,OutputT>	of(DoFn<InputT,OutputT> doFn) Creates a IntraBundleParallelization PTransform for the given DoFn that processes elements using multiple threads.
static IntraBundleParallelization.Unbound	withMaxParallelism(int maxParallelism) Creates a IntraBundleParallelization PTransform with the specified maximum concurrency level.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

IntraBundleParallelization

public IntraBundleParallelization()

Method Detail

of

public static <InputT,OutputT> IntraBundleParallelization.Bound<InputT,OutputT> of(DoFn<InputT,OutputT> doFn)

Creates a IntraBundleParallelization PTransform for the given DoFn that processes elements using multiple threads.

Note that the specified doFn needs to be thread safe.

withMaxParallelism

public static IntraBundleParallelization.Unbound withMaxParallelism(int maxParallelism)

Creates a IntraBundleParallelization PTransform with the specified maximum concurrency level.