如何将List对象转换为JavaDStream Spark?

时间:2016-10-27 04:37:27

标签: java apache-spark spark-streaming sparkcore

我有List对象:

List<Tuple2<Tuple3<String, Long, Object>, Tuple2<Long, Object[]>>> rddQueue =
    Arrays.asList(tupleList);

如何将其转换为JavaDStream Spark? 谢谢提前!

1 个答案:

答案 0 :(得分:4)

您可以使用JavaStreamingContextQueue<T>

来实现这一目标
List<Tuple2<Tuple3<String, Long, Object>, Tuple2<Long, Object[]>>> 
  list = Arrays.asList();

// Random batch time, for the sake of the example.
JavaStreamingContext jssc = 
  new JavaStreamingContext("local[*]", "TestApp", new Duration(4000));

Queue<JavaRDD<Tuple2<Tuple3<String, Long, Object>, Tuple2<Long, Object[]>>>> rddQueue = 
  new LinkedList<>();

rddQueue.add(jssc.sparkContext().parallelize(list));

JavaDStream<Tuple2<Tuple3<String, Long, Object>, Tuple2<Long, Object[]>>> dStream = 
  jssc.queueStream(rddQueue);