Java Code Examples for org.apache.spark.streaming.api.java.JavaPairDStream#mapWithState()

The following examples show how to use org.apache.spark.streaming.api.java.JavaPairDStream#mapWithState() . You can vote up the ones you like or vote down the ones you don't like, and go to the original project or source file by following the links above each example. You may check out the related API usage on the sidebar.
Example 1
Source File: RealtimeTrafficDataProcessor.java    From lambda-arch with Apache License 2.0 5 votes vote down vote up 
/**
 * Method to get total traffic counts of different type of vehicles for each route.
 *
 * @param filteredIotDataStream IoT data stream
 */
public void processTotalTrafficData(JavaDStream<IoTData> filteredIotDataStream) {

    // We need to get count of vehicle group by routeId and vehicleType
    JavaPairDStream<AggregateKey, Long> countDStreamPair = filteredIotDataStream
            .mapToPair(iot -> new Tuple2<>(new AggregateKey(iot.getRouteId(), iot.getVehicleType()), 1L))
            .reduceByKey((a, b) -> a + b);

    // Need to keep state for total count
    StateSpec<AggregateKey, Long, Long, Tuple2<AggregateKey, Long>> stateSpec =
            StateSpec.function(totalSumFunc).timeout(Durations.seconds(3600));

    JavaMapWithStateDStream<AggregateKey, Long, Long, Tuple2<AggregateKey, Long>> countDStreamWithStatePair =
            countDStreamPair.mapWithState(stateSpec);//maintain state for one hour

    // Transform to dstream of TrafficData
    JavaDStream<Tuple2<AggregateKey, Long>> countDStream = countDStreamWithStatePair.map(tuple2 -> tuple2);
    JavaDStream<TotalTrafficData> trafficDStream = countDStream.map(totalTrafficDataFunc);

    // Map Cassandra table column
    Map<String, String> columnNameMappings = new HashMap<String, String>();
    columnNameMappings.put("routeId", "routeid");
    columnNameMappings.put("vehicleType", "vehicletype");
    columnNameMappings.put("totalCount", "totalcount");
    columnNameMappings.put("timeStamp", "timestamp");
    columnNameMappings.put("recordDate", "recorddate");

    // call CassandraStreamingJavaUtil function to save in DB
    javaFunctions(trafficDStream).writerBuilder(
            "traffickeyspace",
            "total_traffic",
            CassandraJavaUtil.mapToRow(TotalTrafficData.class, columnNameMappings)
    ).saveToCassandra();
}
Example 2
Source File: WordCountSocketStateful.java    From Apache-Spark-2x-for-Java-Developers with MIT License 5 votes vote down vote up 
public static void main(String[] args) throws Exception {
 System.setProperty("hadoop.home.dir", "E:\\hadoop");

   SparkConf sparkConf = new SparkConf().setAppName("WordCountSocketEx").setMaster("local[*]");
   JavaStreamingContext streamingContext = new JavaStreamingContext(sparkConf, Durations.seconds(1));
   streamingContext.checkpoint("E:\\hadoop\\checkpoint");
// Initial state RDD input to mapWithState
   @SuppressWarnings("unchecked")
   List<Tuple2<String, Integer>> tuples =Arrays.asList(new Tuple2<>("hello", 1), new Tuple2<>("world", 1));
   JavaPairRDD<String, Integer> initialRDD = streamingContext.sparkContext().parallelizePairs(tuples);
   
   JavaReceiverInputDStream<String> StreamingLines = streamingContext.socketTextStream( "10.0.75.1", Integer.parseInt("9000"), StorageLevels.MEMORY_AND_DISK_SER);
   
   JavaDStream<String> words = StreamingLines.flatMap( str -> Arrays.asList(str.split(" ")).iterator() );
  
   JavaPairDStream<String, Integer> wordCounts = words.mapToPair(str-> new Tuple2<>(str, 1)).reduceByKey((count1,count2) ->count1+count2 );
  


  // Update the cumulative count function
  Function3<String, Optional<Integer>, State<Integer>, Tuple2<String, Integer>> mappingFunc =
      new Function3<String, Optional<Integer>, State<Integer>, Tuple2<String, Integer>>() {
        @Override
        public Tuple2<String, Integer> call(String word, Optional<Integer> one,
            State<Integer> state) {
          int sum = one.orElse(0) + (state.exists() ? state.get() : 0);
          Tuple2<String, Integer> output = new Tuple2<>(word, sum);
          state.update(sum);
          return output;
        }
      };

  // DStream made of get cumulative counts that get updated in every batch
  JavaMapWithStateDStream<String, Integer, Integer, Tuple2<String, Integer>> stateDstream = wordCounts.mapWithState(StateSpec.function(mappingFunc).initialState(initialRDD));

  stateDstream.print();
  streamingContext.start();
  streamingContext.awaitTermination();
}
Example 3
Source File: WordCountRecoverableEx.java    From Apache-Spark-2x-for-Java-Developers with MIT License 5 votes vote down vote up 
protected static JavaStreamingContext createContext(String ip, int port, String checkpointDirectory) {
	SparkConf sparkConf = new SparkConf().setAppName("WordCountRecoverableEx").setMaster("local[*]");
	JavaStreamingContext streamingContext = new JavaStreamingContext(sparkConf, Durations.seconds(1));
	streamingContext.checkpoint(checkpointDirectory);
	// Initial state RDD input to mapWithState
	@SuppressWarnings("unchecked")
	List<Tuple2<String, Integer>> tuples = Arrays.asList(new Tuple2<>("hello", 1), new Tuple2<>("world", 1));
	JavaPairRDD<String, Integer> initialRDD = streamingContext.sparkContext().parallelizePairs(tuples);

	JavaReceiverInputDStream<String> StreamingLines = streamingContext.socketTextStream(ip,port, StorageLevels.MEMORY_AND_DISK_SER);

	JavaDStream<String> words = StreamingLines.flatMap(str -> Arrays.asList(str.split(" ")).iterator());

	JavaPairDStream<String, Integer> wordCounts = words.mapToPair(str -> new Tuple2<>(str, 1))
			.reduceByKey((count1, count2) -> count1 + count2);

	// Update the cumulative count function
	Function3<String, Optional<Integer>, State<Integer>, Tuple2<String, Integer>> mappingFunc = new Function3<String, Optional<Integer>, State<Integer>, Tuple2<String, Integer>>() {
		@Override
		public Tuple2<String, Integer> call(String word, Optional<Integer> one, State<Integer> state) {
			int sum = one.orElse(0) + (state.exists() ? state.get() : 0);
			Tuple2<String, Integer> output = new Tuple2<>(word, sum);
			state.update(sum);
			return output;
		}
	};

	// DStream made of get cumulative counts that get updated in every batch
	JavaMapWithStateDStream<String, Integer, Integer, Tuple2<String, Integer>> stateDstream = wordCounts
			.mapWithState(StateSpec.function(mappingFunc).initialState(initialRDD));

	stateDstream.print();
	return streamingContext;
}
Example 4
Source File: IoTTrafficDataProcessor.java    From iot-traffic-monitor with Apache License 2.0 5 votes vote down vote up 
/**
 * Method to get total traffic counts of different type of vehicles for each route.
 * 
 * @param filteredIotDataStream IoT data stream
 */
public void processTotalTrafficData(JavaDStream<IoTData> filteredIotDataStream) {

	// We need to get count of vehicle group by routeId and vehicleType
	JavaPairDStream<AggregateKey, Long> countDStreamPair = filteredIotDataStream
			.mapToPair(iot -> new Tuple2<>(new AggregateKey(iot.getRouteId(), iot.getVehicleType()), 1L))
			.reduceByKey((a, b) -> a + b);
	
	// Need to keep state for total count
	JavaMapWithStateDStream<AggregateKey, Long, Long, Tuple2<AggregateKey, Long>> countDStreamWithStatePair = countDStreamPair
			.mapWithState(StateSpec.function(totalSumFunc).timeout(Durations.seconds(3600)));//maintain state for one hour

	// Transform to dstream of TrafficData
	JavaDStream<Tuple2<AggregateKey, Long>> countDStream = countDStreamWithStatePair.map(tuple2 -> tuple2);
	JavaDStream<TotalTrafficData> trafficDStream = countDStream.map(totalTrafficDataFunc);

	// Map Cassandra table column
	Map<String, String> columnNameMappings = new HashMap<String, String>();
	columnNameMappings.put("routeId", "routeid");
	columnNameMappings.put("vehicleType", "vehicletype");
	columnNameMappings.put("totalCount", "totalcount");
	columnNameMappings.put("timeStamp", "timestamp");
	columnNameMappings.put("recordDate", "recorddate");

	// call CassandraStreamingJavaUtil function to save in DB
	javaFunctions(trafficDStream).writerBuilder("traffickeyspace", "total_traffic",
			CassandraJavaUtil.mapToRow(TotalTrafficData.class, columnNameMappings)).saveToCassandra();
}
Example 5
Source File: StateFulProcessingExample.java    From Apache-Spark-2x-for-Java-Developers with MIT License 4 votes vote down vote up 
public static void main(String[] args) throws InterruptedException {

		System.setProperty("hadoop.home.dir", "C:\\softwares\\Winutils");

		SparkSession sparkSession = SparkSession.builder().master("local[*]").appName("Stateful Streaming Example")
				.config("spark.sql.warehouse.dir", "file:////C:/Users/sgulati/spark-warehouse").getOrCreate();

		JavaStreamingContext jssc= new JavaStreamingContext(new JavaSparkContext(sparkSession.sparkContext()),
				Durations.milliseconds(1000));
		JavaReceiverInputDStream<String> inStream = jssc.socketTextStream("10.204.136.223", 9999);
		jssc.checkpoint("C:\\Users\\sgulati\\spark-checkpoint");

		JavaDStream<FlightDetails> flightDetailsStream = inStream.map(x -> {
			ObjectMapper mapper = new ObjectMapper();
			return mapper.readValue(x, FlightDetails.class);
		});
		
		

		JavaPairDStream<String, FlightDetails> flightDetailsPairStream = flightDetailsStream
				.mapToPair(f -> new Tuple2<String, FlightDetails>(f.getFlightId(), f));

		Function3<String, Optional<FlightDetails>, State<List<FlightDetails>>, Tuple2<String, Double>> mappingFunc = (
				flightId, curFlightDetail, state) -> {
			List<FlightDetails> details = state.exists() ? state.get() : new ArrayList<>();

			boolean isLanded = false;

			if (curFlightDetail.isPresent()) {
				details.add(curFlightDetail.get());
				if (curFlightDetail.get().isLanded()) {
					isLanded = true;
				}
			}
			Double avgSpeed = details.stream().mapToDouble(f -> f.getTemperature()).average().orElse(0.0);

			if (isLanded) {
				state.remove();
			} else {
				state.update(details);
			}
			return new Tuple2<String, Double>(flightId, avgSpeed);
		};

		JavaMapWithStateDStream<String, FlightDetails, List<FlightDetails>, Tuple2<String, Double>> streamWithState = flightDetailsPairStream
				.mapWithState(StateSpec.function(mappingFunc).timeout(Durations.minutes(5)));
		
		streamWithState.print();
		jssc.start();
		jssc.awaitTermination();
	}
Example 6
Source File: IoTDataProcessor.java    From iot-traffic-monitor with Apache License 2.0 4 votes vote down vote up 
public static void main(String[] args) throws Exception {
 //read Spark and Cassandra properties and create SparkConf
 Properties prop = PropertyFileReader.readPropertyFile();		
 SparkConf conf = new SparkConf()
		 .setAppName(prop.getProperty("com.iot.app.spark.app.name"))
		 .setMaster(prop.getProperty("com.iot.app.spark.master"))
		 .set("spark.cassandra.connection.host", prop.getProperty("com.iot.app.cassandra.host"))
		 .set("spark.cassandra.connection.port", prop.getProperty("com.iot.app.cassandra.port"))
		 .set("spark.cassandra.connection.keep_alive_ms", prop.getProperty("com.iot.app.cassandra.keep_alive"));		 
 //batch interval of 5 seconds for incoming stream		 
 JavaStreamingContext jssc = new JavaStreamingContext(conf, Durations.seconds(5));	
 //add check point directory
 jssc.checkpoint(prop.getProperty("com.iot.app.spark.checkpoint.dir"));
 
 //read and set Kafka properties
 Map<String, String> kafkaParams = new HashMap<String, String>();
 kafkaParams.put("zookeeper.connect", prop.getProperty("com.iot.app.kafka.zookeeper"));
 kafkaParams.put("metadata.broker.list", prop.getProperty("com.iot.app.kafka.brokerlist"));
 String topic = prop.getProperty("com.iot.app.kafka.topic");
 Set<String> topicsSet = new HashSet<String>();
 topicsSet.add(topic);
 //create direct kafka stream
 JavaPairInputDStream<String, IoTData> directKafkaStream = KafkaUtils.createDirectStream(
	        jssc,
	        String.class,
	        IoTData.class,
	        StringDecoder.class,
	        IoTDataDecoder.class,
	        kafkaParams,
	        topicsSet
	    );
 logger.info("Starting Stream Processing");
 
 //We need non filtered stream for poi traffic data calculation
 JavaDStream<IoTData> nonFilteredIotDataStream = directKafkaStream.map(tuple -> tuple._2());
 
 //We need filtered stream for total and traffic data calculation
 JavaPairDStream<String,IoTData> iotDataPairStream = nonFilteredIotDataStream.mapToPair(iot -> new Tuple2<String,IoTData>(iot.getVehicleId(),iot)).reduceByKey((a, b) -> a );

 // Check vehicle Id is already processed
 JavaMapWithStateDStream<String, IoTData, Boolean, Tuple2<IoTData,Boolean>> iotDStreamWithStatePairs = iotDataPairStream
					.mapWithState(StateSpec.function(processedVehicleFunc).timeout(Durations.seconds(3600)));//maintain state for one hour

 // Filter processed vehicle ids and keep un-processed
 JavaDStream<Tuple2<IoTData,Boolean>> filteredIotDStreams = iotDStreamWithStatePairs.map(tuple2 -> tuple2)
					.filter(tuple -> tuple._2.equals(Boolean.FALSE));

 // Get stream of IoTdata
 JavaDStream<IoTData> filteredIotDataStream = filteredIotDStreams.map(tuple -> tuple._1);
 
 //cache stream as it is used in total and window based computation
 filteredIotDataStream.cache();
 	 
 //process data
 IoTTrafficDataProcessor iotTrafficProcessor = new IoTTrafficDataProcessor();
 iotTrafficProcessor.processTotalTrafficData(filteredIotDataStream);
 iotTrafficProcessor.processWindowTrafficData(filteredIotDataStream);

 //poi data
 POIData poiData = new POIData();
 poiData.setLatitude(33.877495);
 poiData.setLongitude(-95.50238);
 poiData.setRadius(30);//30 km
 
 //broadcast variables. We will monitor vehicles on Route 37 which are of type Truck
 Broadcast<Tuple3<POIData, String, String>> broadcastPOIValues = jssc.sparkContext().broadcast(new Tuple3<>(poiData,"Route-37","Truck"));
 //call method  to process stream
 iotTrafficProcessor.processPOIData(nonFilteredIotDataStream,broadcastPOIValues);
 
 //start context
 jssc.start();            
 jssc.awaitTermination();  
}
Example 7
Source File: WordCountingAppWithCheckpoint.java    From tutorials with MIT License 4 votes vote down vote up 
public static void main(String[] args) throws InterruptedException {

        Logger.getLogger("org")
            .setLevel(Level.OFF);
        Logger.getLogger("akka")
            .setLevel(Level.OFF);

        Map<String, Object> kafkaParams = new HashMap<>();
        kafkaParams.put("bootstrap.servers", "localhost:9092");
        kafkaParams.put("key.deserializer", StringDeserializer.class);
        kafkaParams.put("value.deserializer", StringDeserializer.class);
        kafkaParams.put("group.id", "use_a_separate_group_id_for_each_stream");
        kafkaParams.put("auto.offset.reset", "latest");
        kafkaParams.put("enable.auto.commit", false);

        Collection<String> topics = Arrays.asList("messages");

        SparkConf sparkConf = new SparkConf();
        sparkConf.setMaster("local[2]");
        sparkConf.setAppName("WordCountingAppWithCheckpoint");
        sparkConf.set("spark.cassandra.connection.host", "127.0.0.1");

        JavaStreamingContext streamingContext = new JavaStreamingContext(sparkConf, Durations.seconds(1));

        sparkContext = streamingContext.sparkContext();

        streamingContext.checkpoint("./.checkpoint");

        JavaInputDStream<ConsumerRecord<String, String>> messages = KafkaUtils.createDirectStream(streamingContext, LocationStrategies.PreferConsistent(), ConsumerStrategies.<String, String> Subscribe(topics, kafkaParams));

        JavaPairDStream<String, String> results = messages.mapToPair(record -> new Tuple2<>(record.key(), record.value()));

        JavaDStream<String> lines = results.map(tuple2 -> tuple2._2());

        JavaDStream<String> words = lines.flatMap(x -> Arrays.asList(x.split("\\s+"))
            .iterator());

        JavaPairDStream<String, Integer> wordCounts = words.mapToPair(s -> new Tuple2<>(s, 1))
            .reduceByKey((Function2<Integer, Integer, Integer>) (i1, i2) -> i1 + i2);

        JavaMapWithStateDStream<String, Integer, Integer, Tuple2<String, Integer>> cumulativeWordCounts = wordCounts.mapWithState(StateSpec.function((word, one, state) -> {
            int sum = one.orElse(0) + (state.exists() ? state.get() : 0);
            Tuple2<String, Integer> output = new Tuple2<>(word, sum);
            state.update(sum);
            return output;
        }));

        cumulativeWordCounts.foreachRDD(javaRdd -> {
            List<Tuple2<String, Integer>> wordCountList = javaRdd.collect();
            for (Tuple2<String, Integer> tuple : wordCountList) {
                List<Word> wordList = Arrays.asList(new Word(tuple._1, tuple._2));
                JavaRDD<Word> rdd = sparkContext.parallelize(wordList);
                javaFunctions(rdd).writerBuilder("vocabulary", "words", mapToRow(Word.class))
                    .saveToCassandra();
            }
        });

        streamingContext.start();
        streamingContext.awaitTermination();
    }