Downloaded 113 times
![Nested parallelStreams
Outer parallelStream could exhaust ForkJoin
Workers:
long start = System.currentTimeMillis();
IntStream.range(0, 10_000).parallel()
.forEach(i -> {
results[i][0] = (int) Math.round(Math.random() * 100);
IntStream.range(1, 9_999)
.parallel().forEach((int j) ->
results[i][j] =
(int) Math.round(Math.random() * 1000));
});
Process finalized in 22974 msecs
Process finalized in 22575 msecs
Process finalized in 22606 msecs](https://image.slidesharecdn.com/parallelstreamsenjava8-150305053752-conversion-gate01/75/Parallel-streams-in-java-8-35-2048.jpg)
![Nested parallelStreams
Outer parallelStream could exhaust ForkJoin
Workers:
long start = System.currentTimeMillis();
IntStream.range(0, 10_000).parallel()
.forEach(i -> {
results[i][0] = (int) Math.round(Math.random() * 100);
IntStream.range(1, 9_999)
.sequential().forEach((int j) ->
results[i][j] =
(int) Math.round(Math.random() * 1000));
});
Process finalized in 12491 msecs
Process finalized in 12589 msecs
Process finalized in 12798 msecs](https://image.slidesharecdn.com/parallelstreamsenjava8-150305053752-conversion-gate01/75/Parallel-streams-in-java-8-36-2048.jpg)
Uploaded byDavid Gómez García
Parallel streams in java 8
The document discusses the use of parallel streams in Java 8 for concurrent data processing, emphasizing their efficiency in handling large datasets through the fork/join framework. It provides examples of filtering and sorting operations, compares performance between sequential and parallel streams, and addresses potential pitfalls such as blocking I/O and excessive autoboxing. Overall, it advises on best practices to optimize the use of parallel streams while cautioning against common misuse.
More Related Content
Similar to Parallel streams in java 8
Parallel streams in java 8
- 1. ParallelStreams Concurrent data processingin Java 8 David Gómez G. @dgomezg dgomezg@autentia.com
- 2. Do you remember? usestream() for (int i = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, Thread.activeCount()); } 4999299 elements computed in 225 msecs with 9 threads 4999299 elements computed in 230 msecs with 9 threads 4999299 elements computed in 250 msecs with 9 threads @dgomezg
- 3.
- 4.
- 5. A Stream is… Anconvenience method to iterate over collections in a declarative way List<Integer> numbers = new ArrayList<Integer>(); for (int i= 0; i < 100 ; i++) { numbers.add(i); } List<Integer> evenNumbers = numbers.stream() .filter(n -> n % 2 == 0) .collect(toList()); @dgomezg
- 6. Anatomy of aStream Source Intermediate Operations filter map order function Final operation pipeline @dgomezg
- 7. Iterating a Stream List<Integer>evenNumbers = numbers.stream() .filter(n -> n % 2 == 0) .collect(toList()); Internal Iteration - No manual Iterators handling - Concise - Fluent API: chain sequence processing Elements computed only when needed @dgomezg
- 8. Iterating a Stream List<Integer>evenNumbers = numbers.parallelStream() .filter(n -> n % 2 == 0) .collect(toList()); Easily Parallelism - Concurrency is hard to be done right! - Uses ForkJoin - Process steps should be - stateless - independent @dgomezg
- 9. Parallel Streams use stream() List<Integer>numbers = new ArrayList<>(); for (int i= 0; i < 10_000_000 ; i++) { numbers.add((int)Math.round(Math.random()*100)); } //This will use just a single thread Stream<Integer> evenNumbers = numbers.stream(); or parallelStream() //Automatically select the optimum number of threads Stream<Integer> evenNumbers = numbers.parallelStream(); @dgomezg
- 10. Let’s test it usestream() for (int i = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.stream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, Thread.activeCount()); } 5001983 elements computed in 828 msecs with 2 threads 5001983 elements computed in 843 msecs with 2 threads 5001983 elements computed in 675 msecs with 2 threads 5001983 elements computed in 795 msecs with 2 threads @dgomezg
- 11. Going parallel use stream() for(int i = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, Thread.activeCount()); } 4999299 elements computed in 225 msecs with 9 threads 4999299 elements computed in 230 msecs with 9 threads 4999299 elements computed in 250 msecs with 9 threads @dgomezg
- 12.
- 13.
- 14. Fork/Join Framework Proposed byDoug Lea "a style of parallel programming in which problems are solved by (recursively) splitting them into subtasks that are solved in parallel." Available in Java 7 Used by ParallelStreams
- 15. The F/J algorithm Resultsolve(Problem problem) { if (problem is small) directly solve problem else { split problem into independent parts fork new subtasks to solve each part join all subtasks compose result from subresults } } as proposed by Doug Lea
- 16. ForkJoinPool ExecutorService implementation that •has a defined number of Workers (threads) • executes ForkJoinTasks • submitted by execute(ForkJoinTask task) • or by invoke(ForkJoinTask task)
- 17. ForkJoinTask Abstract class thatrepresents a task to be run concurrently Every ForkJoinTask could be splitted (if not small enough) and solved Recursively Two concrete implementations • RecursiveAction if not returning value • RecursiveTask if returning a value
- 18. ForkJoinWorkerThread Any of thethreads created by the ForkJoinPool Executes ForkJoinTasks Everyone has a Dequeue for tasks (allows task stealing)
- 19. ForkJoinWorkerThread Result solve(Problem problem) { if(problem is small) directly solve problem else { split problem into independent parts fork new subtasks to solve each part join all subtasks compose result from subresults } } the F/J algorithm plus Task Stealing.
- 20. Fork/Join. When touse? For computations that could be splitted into smaller tasks aka ‘divide and conquer’ algorithms Independent Reduction with no contention.
- 21.
- 22. ParallellStreams for (int i= 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, Thread.activeCount()); } 4999299 elements computed in 225 msecs with 9 threads 4999299 elements computed in 230 msecs with 9 threads 4999299 elements computed in 250 msecs with 9 threads
- 23. Thread.activeCount not accurate for(int i = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, Thread.activeCount()); } Thread.activeCount() does not show the effective number of threads processing the stream
- 24. Better count threadsinvolved Set<String> workerThreadNames = new ConcurrentSet<>(); for (int i = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.stream() .filter(n -> n % 2 == 0) .peek(n -> workerThreadNames.add( Thread.currentThread().getName())) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, workerThreadNames.size()); }
- 25. Threads usage ParallelStreams usethe common ForkJoinPool Number of worker threads configured with -‐Djava.util.concurrent.ForkJoinPool.common.parallelism=n Useful to keep CPU parallelism under control… …but …
- 26. Limiting parallelism for (inti = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.stream() .filter(n -> n % 2 == 0) .peek(n -> workerThreadNames.add( Thread.currentThread().getName())) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, workerThreadNames.size()); } -‐Djava.util.concurrent.ForkJoinPool.common.parallelism=4 5001069 elements computed in 269 msecs with 5 threads WTF
- 27. Limiting parallelism for (inti = 0; i < 100; i++) { long start = System.currentTimeMillis(); List<Integer> even = numbers.stream() .filter(n -> n % 2 == 0) .peek(n -> workerThreadNames.add( Thread.currentThread().getName())) .sorted() .collect(toList()); System.out.printf( "%d elements computed in %5d msecs with %d threadsn”, even.size(), System.currentTimeMillis() - start, workerThreadNames.size()); } System.out.println("credits to threads: “ + workerThreadNames); 5001069 elements computed in 269 msecs with 5 threads credits to threads: ForkJoinPool.commonPool-worker-0, ForkJoinPool.commonPool-worker-1, ForkJoinPool.commonPool-worker-2, ForkJoinPool.commonPool-worker-3, main WTF
- 28. Threads Involved inParallelStream ParallelStreams use the common ForkJoinPool Thread invoking ParallelStream also used as Worker Caveats: •ParallelStream processing is synchronous for invoking thread •Other Threads using common ForkJoinPool could be affected
- 29. ParallelStream Hack ParallelStream canbe forced to use a custom ForkJoinPool ForkJoinPool forkJoinPool = new ForkJoinPool(4); long start = System.currentTimeMillis(); numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList());
- 30. ParallelStream Hack ParallelStream canbe forced to use a custom ForkJoinPool ForkJoinPool forkJoinPool = new ForkJoinPool(4); long start = System.currentTimeMillis(); ForkJoinTask<List<Integer>> task = forkJoinPool.submit(() -> { return numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); } ); List<Integer> even = task.get();
- 31. ParallelStream Hack ParallelStream canbe forced to use a custom ForkJoinPool ForkJoinPool forkJoinPool = new ForkJoinPool(4); ForkJoinTask<List<Integer>> task = forkJoinPool.submit(() -> { return numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); } ); List<Integer> even = task.get(); Task submitted in 1 msecs 5000805 elements computed in 328 msecs with 4 threads
- 32. ParallelStream Hack benefits Acustom ExecutorService • Does not affect other ParallelStreams • Does not affect Common ForkJoinPool users • Reduces unpredictable latency due to other CommonForkJoin Pool load • Invoking thread not used as worker (async parallel process)
- 33. Problems derived from CommonForkJoinPool
- 34. Blocking for IO Iffirsts URLs stuck on a ConnectionTimeOut, overall performance could be affected Stream<String> urls = Files.lines(Paths.get("urlsToCheck.txt")); List<String> errors = urls.parallel().filter(url -> { //Connect to URL and wait for 200 response or timeout return true; }).collect(toList());
- 35. Nested parallelStreams Outer parallelStreamcould exhaust ForkJoin Workers: long start = System.currentTimeMillis(); IntStream.range(0, 10_000).parallel() .forEach(i -> { results[i][0] = (int) Math.round(Math.random() * 100); IntStream.range(1, 9_999) .parallel().forEach((int j) -> results[i][j] = (int) Math.round(Math.random() * 1000)); }); Process finalized in 22974 msecs Process finalized in 22575 msecs Process finalized in 22606 msecs
- 36. Nested parallelStreams Outer parallelStreamcould exhaust ForkJoin Workers: long start = System.currentTimeMillis(); IntStream.range(0, 10_000).parallel() .forEach(i -> { results[i][0] = (int) Math.round(Math.random() * 100); IntStream.range(1, 9_999) .sequential().forEach((int j) -> results[i][j] = (int) Math.round(Math.random() * 1000)); }); Process finalized in 12491 msecs Process finalized in 12589 msecs Process finalized in 12798 msecs
- 37.
- 38. Too much Auto(un)boxing outboxingand boxing of Integers in every filter call List<Integer> even = numbers.parallelStream() .filter(n -> n % 2 == 0) .sorted() .collect(toList()); 4999464 elements computed in 290 msecs with 8 threads 4999464 elements computed in 276 msecs with 8 threads 4999464 elements computed in 257 msecs with 8 threads 4999464 elements computed in 265 msecs with 8 threads
- 39. Less Auto(un)boxing outboxing andboxing of Integers in every filter call List<Integer> even = numbers.parallelStream() .mapToInt(n -> n) .filter(n -> n % 2 == 0) .sorted() .boxed() .collect(toList()); 4999460 elements computed in 160 msecs with 8 threads 4999460 elements computed in 243 msecs with 8 threads 4999460 elements computed in 144 msecs with 8 threads 4999460 elements computed in 140 msecs with 8 threads
- 40.
- 41. Conclusions ParallelStreams eases concurrentprocessing but: • Understand how it works • Don’t abuse the default common ForkJoinPool • Don’t use when blocking by IO • Or use a custom ForkJoinPool • Avoid unnecessary autoboxing • Don’t add contention or synchronisation • Be careful with nested parallel streams • Use method references when sorting
- 42.