浅谈Spark中的宽依赖和窄依赖

1、何为Spark中的宽依赖和窄依赖

1.1、官方源码解释

1.1.1、NarrowDependency（窄依赖）

/**
 * :: DeveloperApi ::
 * Base class for dependencies where each partition of the child RDD depends on a small number
 * of partitions of the parent RDD. Narrow dependencies allow for pipelined execution.
 */
@DeveloperApi
abstract class NarrowDependency[T](_rdd: RDD[T]) extends Dependency[T] {
  /**
   * Get the parent partitions for a child partition.
   * @param partitionId a partition of the child RDD
   * @return the partitions of the parent RDD that the child partition depends upon
   */
  def getParents(partitionId: Int): Seq[Int]

  override def rdd: RDD[T] = _rdd
}

如上源码注释所讲，这些最基本的依赖关系，其中子RDD的每个分区依赖于父RDD的少量分区，并且窄依赖允许通过一个管道来执行。

总结来说：父RDD的一个分区只会被子RDD的1个分区所依赖，并且其中是不会产生shuffle过程的。

下图可以帮助理解

1.1.2、ShuffleDependency（宽依赖）

/**
 * :: DeveloperApi ::
 * Represents a dependency on the output of a shuffle stage. Note that in the case of shuffle,
 * the RDD is transient since we don't need it on the executor side.
 *
 * @param _rdd the parent RDD
 * @param partitioner partitioner used to partition the shuffle output
 * @param serializer [[org.apache.spark.serializer.Serializer Serializer]] to use. If not set
 *                   explicitly then the default serializer, as specified by `spark.serializer`
 *                   config option, will be used.
 * @param keyOrdering key ordering for RDD's shuffles
 * @param aggregator map/reduce-side aggregator for RDD's shuffle
 * @param mapSideCombine whether to perform partial aggregation (also known as map-side combine)
 */
@DeveloperApi
class ShuffleDependency[K: ClassTag, V: ClassTag, C: ClassTag](
    @transient private val _rdd: RDD[_ <: Product2[K, V]],
    val partitioner: Partitioner,
    val serializer: Serializer = SparkEnv.get.serializer,
    val keyOrdering: Option[Ordering[K]] = None,
    val aggregator: Option[Aggregator[K, V, C]] = None,
    val mapSideCombine: Boolean = false)
  extends Dependency[Product2[K, V]] {

  if (mapSideCombine) {
    require(aggregator.isDefined, "Map-side combine without Aggregator specified!")
  }
  override def rdd: RDD[Product2[K, V]] = _rdd.asInstanceOf[RDD[Product2[K, V]]]

  private[spark] val keyClassName: String = reflect.classTag[K].runtimeClass.getName
  private[spark] val valueClassName: String = reflect.classTag[V].runtimeClass.getName
  // Note: It's possible that the combiner class tag is null, if the combineByKey
  // methods in PairRDDFunctions are used instead of combineByKeyWithClassTag.
  private[spark] val combinerClassName: Option[String] =
    Option(reflect.classTag[C]).map(_.runtimeClass.getName)

  val shuffleId: Int = _rdd.context.newShuffleId()

  val shuffleHandle: ShuffleHandle = _rdd.context.env.shuffleManager.registerShuffle(
    shuffleId, _rdd.partitions.length, this)

  _rdd.sparkContext.cleaner.foreach(_.registerShuffleForCleanup(this))
}

源码注释可以理解为：表示对shuffle阶段的输出的依赖关系。注意，在shuffle的情况下，RDD是暂时的，因为我们不需要在executor端使用它。其实看dependency名字就知道了，顾名思义，只有发生了shuffle才可以称之为宽依赖

可以总结如下：父RDD的一个分区会被子RDD的多个分区所依赖

下图可以帮助理解

2、为什么需要宽窄依赖

2.1、DAG和Stage概念

2.1.1、DAG

Spark的DAG:就是spark任务/程序执行的流程图，

DAG的开始:从创建RDD开始

DAG的结束:到Action结束

一个Spark程序中有几个Action操作就有几个DAG，如何区分算子是否是action还是transformation，如果api返回的是RDD，则这个api算子一定是transformation，反之则为action。

2.1.2、Stage

Stage:是DAG中根据shuffle划分出来的阶段，前面的阶段执行完才可以执行后面的阶段，同一个阶段中的各个任务可以并行执行无需等待。

2.2、详解如何宽窄依赖之间是如何切分成Stage的、

如图：

总结:

窄依赖: 并行化+容错

宽依赖: 进行阶段划分(shuffle后的阶段需要等待shuffle前的阶段计算完才能执行)

3、哪些算子是宽依赖那些算子是窄依赖

map、flatMap、filter等等常规情况下都是窄依赖，不会产生shuffle

reduceBykey、groupByKey等等常规情况下都是宽依赖，会产生shuffle

但是其实宽窄依赖不能光靠算子名称来划分，需要根据定义来划分，因为有的时候，join操作可能是窄依赖，有的时候就是宽依赖。具体的还是要看stage具体执行的时候划分。

4、根据Spark UI解读join何时是窄依赖何时是宽依赖

4.1、join窄依赖

当join和前一个父RDD被划分到同一个Stage中的时候，就可以认为这是一个窄依赖

        SparkConf conf = new SparkConf().setAppName("Java-Test-WordCount").setMaster("local[*]");
        JavaSparkContext jsc = new JavaSparkContext(conf);
        List<Tuple2<String, Integer>> tuple2List1 = Arrays.asList(new Tuple2<>("Alice", 15), new Tuple2<>("Bob", 18), new Tuple2<>("Thomas", 20), new Tuple2<>("Catalina", 25));
        List<Tuple3<String, String, String>> tuple3List = Arrays.asList(new Tuple3<>("Alice", "Female", "NanJ"), new Tuple3<>("Thomas", "Male", "ShangH"), new Tuple3<>("Tom", "Male", "BeiJ"));

		//通过parallelize构建第一个RDD
        JavaRDD<Tuple2<String, Integer>> javaRDD1 = jsc.parallelize(tuple2List1);
		
		//通过parallelize构建第二个RDD
        JavaRDD<Tuple3<String, String, String>> javaRDD2 = jsc.parallelize(tuple3List);

		//通过mapToPair根据第一个RDD构建第三个RDD
        JavaPairRDD<String, Integer> javaRDD3 = javaRDD1.mapToPair(new PairFunction<Tuple2<String, Integer>, String, Integer>() {
            @Override
            public Tuple2<String, Integer> call(Tuple2<String, Integer> tuple2) {
                return tuple2;
            }
        });

		//通过partitionBy根据第三个RDD构建第五个RDD
        JavaPairRDD<String, Integer> javaRDD31 = javaRDD3.partitionBy(new HashPartitioner(2));

		//通过mapToPair根据第二个RDD构建第四个RDD
        JavaPairRDD<String, Tuple2<String, String>> javaRDD4 = javaRDD2.mapToPair(new PairFunction<Tuple3<String, String, String>, String, Tuple2<String, String>>() {
            @Override
            public Tuple2<String, Tuple2<String, String>> call(Tuple3<String, String, String> tuple3) {
                return new Tuple2<>(tuple3._1(), new Tuple2<>(tuple3._2(), tuple3._3()));
            }
        });

		//通过partitionBy根据第四个RDD构建第六个RDD
        JavaPairRDD<String, Tuple2<String, String>> javaRDD41 = javaRDD4.partitionBy(new HashPartitioner(2));

		//通过join 根据第五和第六个RDD构建出第七个RDD
        JavaPairRDD<String, Tuple2<Integer, Tuple2<String, String>>> javaRDD6 = javaRDD31.join(javaRDD41);

        javaRDD6.foreach(new VoidFunction<Tuple2<String, Tuple2<Integer, Tuple2<String, String>>>>() {
            @Override
            public void call(Tuple2<String, Tuple2<Integer, Tuple2<String, String>>> stringTuple2Tuple2) throws Exception {
                System.out.print(stringTuple2Tuple2);
            }
        });

查看Spark Web UI

Note：通过如上UI显示，可以看出，Stage5中，partitionBy 和 join在同一个Stage中，并且join是子RDD的算子，故而可以得出结论，在此Stage中，join就是一个窄依赖

4.2、join宽依赖

        SparkConf conf = new SparkConf().setAppName("Java-Test-WordCount").setMaster("local[*]");
        JavaSparkContext jsc = new JavaSparkContext(conf);
        List<Tuple2<String, Integer>> tuple2List1 = Arrays.asList(new Tuple2<>("Alice", 15), new Tuple2<>("Bob", 18), new Tuple2<>("Thomas", 20), new Tuple2<>("Catalina", 25));
        List<Tuple3<String, String, String>> tuple3List = Arrays.asList(new Tuple3<>("Alice", "Female", "NanJ"), new Tuple3<>("Thomas", "Male", "ShangH"), new Tuple3<>("Tom", "Male", "BeiJ"));

		//通过parallelize构建第一个RDD
        JavaRDD<Tuple2<String, Integer>> javaRDD1 = jsc.parallelize(tuple2List1);
		
		//通过parallelize构建第二个RDD
        JavaRDD<Tuple3<String, String, String>> javaRDD2 = jsc.parallelize(tuple3List);

		//通过mapToPair根据第一个RDD构建第三个RDD
        JavaPairRDD<String, Integer> javaRDD3 = javaRDD1.mapToPair(new PairFunction<Tuple2<String, Integer>, String, Integer>() {
            @Override
            public Tuple2<String, Integer> call(Tuple2<String, Integer> tuple2) {
                return tuple2;
            }
        });


		//通过mapToPair根据第二个RDD构建第四个RDD
        JavaPairRDD<String, Tuple2<String, String>> javaRDD4 = javaRDD2.mapToPair(new PairFunction<Tuple3<String, String, String>, String, Tuple2<String, String>>() {
            @Override
            public Tuple2<String, Tuple2<String, String>> call(Tuple3<String, String, String> tuple3) {
                return new Tuple2<>(tuple3._1(), new Tuple2<>(tuple3._2(), tuple3._3()));
            }
        });

		//通过join 根据第三个RDD和第四个RDD构建得出第五个RDD
		JavaPairRDD<String, Tuple2<Integer, Tuple2<String, String>>> javaRDD5 = javaRDD3.join(javaRDD4);

        javaRDD5.foreach(new VoidFunction<Tuple2<String, Tuple2<Integer, Tuple2<String, String>>>>() {
            @Override
            public void call(Tuple2<String, Tuple2<Integer, Tuple2<String, String>>> stringTuple2Tuple2) throws Exception {
                System.out.print(stringTuple2Tuple2);
            }
        });

查看Spark Web UI

Note：通过如上UI显示，可以看出Stage2中就只有join一个算子操作，故而此处的join算子就是一个宽依赖。