Examples with Fields backtype.storm.tuple.Fields used on opensource projects

Example 31 with Fields

use of backtype.storm.tuple.Fields in project jstorm by alibaba.

the class SequenceTopologyTest method testSequenceTopology.

@Test
public void testSequenceTopology() {
    TopologyBuilder topologyBuilder = new TopologyBuilder();
    topologyBuilder.setSpout(SequenceTopologyDef.SEQUENCE_SPOUT_NAME, new SequenceTestSpout(), SPOUT_PARALLELISM_HINT);
    topologyBuilder.setBolt(SequenceTopologyDef.SPLIT_BOLT_NAME, new SequenceTestSplitRecord(), BOLT_PARALLELISM_HINT).localOrShuffleGrouping(SequenceTopologyDef.SEQUENCE_SPOUT_NAME);
    topologyBuilder.setBolt(SequenceTopologyDef.TRADE_BOLT_NAME, new SequenceTestPairCount(), BOLT_PARALLELISM_HINT).shuffleGrouping(SequenceTopologyDef.SPLIT_BOLT_NAME, SequenceTopologyDef.TRADE_STREAM_ID);
    topologyBuilder.setBolt(SequenceTopologyDef.CUSTOMER_BOLT_NAME, new SequenceTestPairCount(), BOLT_PARALLELISM_HINT).shuffleGrouping(SequenceTopologyDef.SPLIT_BOLT_NAME, SequenceTopologyDef.CUSTOMER_STREAM_ID);
    topologyBuilder.setBolt(SequenceTopologyDef.MERGE_BOLT_NAME, new SequenceTestMergeRecord(), BOLT_PARALLELISM_HINT).fieldsGrouping(SequenceTopologyDef.TRADE_BOLT_NAME, new Fields("ID")).fieldsGrouping(SequenceTopologyDef.CUSTOMER_BOLT_NAME, new Fields("ID"));
    topologyBuilder.setBolt(SequenceTopologyDef.TOTAL_BOLT_NAME, new SequenceTestTotalCount(), BOLT_PARALLELISM_HINT).noneGrouping(SequenceTopologyDef.MERGE_BOLT_NAME);
    // use config in detail.yaml
    Map conf = new HashMap();
    // Config.setFallBackOnJavaSerialization(conf, true);      //fall.back.on.java.serialization: true
    // //enable.split: true
    // Config.registerSerialization(conf, TradeCustomer.class, TradeCustomerSerializer.class);
    // Config.registerSerialization(conf, Pair.class, PairSerializer.class);
    Config.setNumAckers(conf, 1);
    Config.setNumWorkers(conf, 3);
    // set a limit for the spout to get a precise
    conf.put("spout.max.sending.num", SPOUT_MAX_SEND_NUM);
    // number to make sure the topology works well.
    conf.put(Config.TOPOLOGY_NAME, "SequenceTopologyTest");
    // the following is just for the JStormUnitTestMetricValidator to pick the metric data
    // from all the metrics.If you are not using JStormUnitTestMetricValidator, it is useless.
    // The first element is the key that register in the metric, the second one is the key
    // map with the metric value as a parameter in the callback function validateMetrics().
    Set<String> userDefineMetrics = new HashSet<String>();
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_SPOUT_EMIT);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_SPOUT_SUCCESS);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_SPOUT_FAIL);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_SPOUT_TRADE_SUM);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_SPOUT_CUSTOMER_SUM);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_SPLIT_EMIT);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_PAIR_TRADE_EMIT);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_PAIR_CUSTOMER_EMIT);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_MERGE_EMIT);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_TOTAL_EXECUTE);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_TOTAL_TRADE_SUM);
    userDefineMetrics.add(SequenceTestMetricsDef.METRIC_TOTAL_CUSTOMER_SUM);
    JStormUnitTestMetricValidator validator = new JStormUnitTestMetricValidator(userDefineMetrics) {

        @Override
        public boolean validateMetrics(Map<String, Double> metrics) {
            for (Map.Entry<String, Double> entry : metrics.entrySet()) LOG.info("user define metric Key = " + entry.getKey() + " Value = " + entry.getValue());
            int spoutEmit = (int) (metrics.get(SequenceTestMetricsDef.METRIC_SPOUT_EMIT)).doubleValue();
            int spoutSuccess = (int) (metrics.get(SequenceTestMetricsDef.METRIC_SPOUT_SUCCESS)).doubleValue();
            int spoutFail = (int) (metrics.get(SequenceTestMetricsDef.METRIC_SPOUT_FAIL)).doubleValue();
            long spoutTradeSum = (long) (metrics.get(SequenceTestMetricsDef.METRIC_SPOUT_TRADE_SUM)).doubleValue();
            long spoutCustomerSum = (long) (metrics.get(SequenceTestMetricsDef.METRIC_SPOUT_CUSTOMER_SUM)).doubleValue();
            int splitEmit = (int) (metrics.get(SequenceTestMetricsDef.METRIC_SPLIT_EMIT)).doubleValue();
            int pairTradeEmit = (int) (metrics.get(SequenceTestMetricsDef.METRIC_PAIR_TRADE_EMIT)).doubleValue();
            int pairCustomerEmit = (int) (metrics.get(SequenceTestMetricsDef.METRIC_PAIR_CUSTOMER_EMIT)).doubleValue();
            int mergeEmit = (int) (metrics.get(SequenceTestMetricsDef.METRIC_MERGE_EMIT)).doubleValue();
            int totalExecute = (int) (metrics.get(SequenceTestMetricsDef.METRIC_TOTAL_EXECUTE)).doubleValue();
            long totalTradeSum = (long) (metrics.get(SequenceTestMetricsDef.METRIC_TOTAL_TRADE_SUM)).doubleValue();
            long totalCustomerSum = (long) (metrics.get(SequenceTestMetricsDef.METRIC_TOTAL_CUSTOMER_SUM)).doubleValue();
            assertEquals(SPOUT_MAX_SEND_NUM, spoutEmit);
            assertEquals(spoutEmit, spoutSuccess);
            assertEquals(0, spoutFail);
            assertEquals(2 * spoutEmit, splitEmit);
            assertEquals(splitEmit, pairTradeEmit * 2);
            assertEquals(splitEmit, pairCustomerEmit * 2);
            assertEquals(splitEmit, mergeEmit * 2);
            assertEquals(mergeEmit, totalExecute);
            assertEquals(spoutTradeSum, totalTradeSum);
            assertEquals(spoutCustomerSum, totalCustomerSum);
            return true;
        }
    };
    // the below line time in second 150 is recommend, at least it should be more than 120 since the
    // metric data was grabbed every 60s but not so precise.
    boolean result = JStormUnitTestRunner.submitTopology(topologyBuilder.createTopology(), conf, 150, validator);
    assertTrue("Topology should pass the validator", result);
}

Example 32 with Fields

use of backtype.storm.tuple.Fields in project jstorm by alibaba.

the class TridentMinMaxOfVehiclesTest method testTridentMinMaxOfVehicles.

@Test
public void testTridentMinMaxOfVehicles() {
    Fields driverField = new Fields(Driver.FIELD_NAME);
    Fields vehicleField = new Fields(Vehicle.FIELD_NAME);
    Fields fields = new Fields(Vehicle.FIELD_NAME, Driver.FIELD_NAME);
    Random random = new Random(System.currentTimeMillis());
    List<Values> vehicleContent = new ArrayList<Values>();
    List<Values> driverContent = new ArrayList<Values>();
    int maxSpeed = -1, minSpeed = 10000;
    double maxEfficiency = -1, minEfficiency = 10000;
    for (int i = 0; i < SPOUT_BATCH_SIZE; i++) {
        int speed = random.nextInt(10000);
        maxSpeed = Math.max(speed, maxSpeed);
        minSpeed = Math.min(speed, minSpeed);
        double efficiency = random.nextDouble() * 10000;
        maxEfficiency = Math.max(efficiency, maxEfficiency);
        minEfficiency = Math.min(efficiency, minEfficiency);
        vehicleContent.add(new Values(new Vehicle("vehicle-" + (i + 1), speed, efficiency)));
        driverContent.add(new Values(new Driver("driver-" + (i + 1), i + 1)));
    }
    ShuffleValuesBatchSpout spout = new ShuffleValuesBatchSpout(fields, vehicleContent, driverContent);
    TridentTopology tridentTopology = new TridentTopology();
    Stream vehiclesStream = tridentTopology.newStream("spout", spout).each(fields, new Debug("#### vehicles"));
    Stream slowVehiclesStream = vehiclesStream.min(new SpeedComparator()).each(vehicleField, new Debug("#### slowest vehicle")).peek(new SpeedValidator(minSpeed));
    Stream slowDriversStream = slowVehiclesStream.project(driverField).each(driverField, new Debug("#### slowest driver"));
    vehiclesStream.max(new SpeedComparator()).each(vehicleField, new Debug("#### fastest vehicle")).peek(new SpeedValidator(maxSpeed)).project(driverField).each(driverField, new Debug("#### fastest driver"));
    vehiclesStream.minBy(Vehicle.FIELD_NAME, new EfficiencyComparator()).each(vehicleField, new Debug("#### least efficient vehicle")).peek(new EfficiencyValidator(minEfficiency));
    vehiclesStream.maxBy(Vehicle.FIELD_NAME, new EfficiencyComparator()).each(vehicleField, new Debug("#### most efficient vehicle")).peek(new EfficiencyValidator(maxEfficiency));
    Map config = new HashMap();
    config.put(Config.TOPOLOGY_NAME, "TridentMinMaxOfVehiclesTest");
    // use the assert in the body of consumer.accept() to validate
    JStormUnitTestRunner.submitTopology(tridentTopology.build(), null, 120, null);
}

Example 33 with Fields

use of backtype.storm.tuple.Fields in project jstorm by alibaba.

the class TridentTumblingCountWindowTest method testTridentTumblingCountWindow.

@Test
public void testTridentTumblingCountWindow() {
    WindowsStoreFactory windowsStoreFactory = new InMemoryWindowsStoreFactory();
    FixedLimitBatchSpout spout = new FixedLimitBatchSpout(SPOUT_LIMIT, new Fields("sentence"), SPOUT_BATCH_SIZE, new Values("the cow jumped over the moon"), new Values("the man went to the store and bought some candy"), new Values("four score and seven years ago"), new Values("how many apples can you eat"), new Values("to be or not to be the person"));
    TridentTopology tridentTopology = new TridentTopology();
    Stream stream = tridentTopology.newStream("spout1", spout).parallelismHint(16).each(new Fields("sentence"), new Split(), new Fields("word")).window(windowConfig, windowsStoreFactory, new Fields("word"), new CountAsAggregator(), new Fields("count")).peek(new ValidateConsumer());
    Map config = new HashMap();
    config.put(Config.TOPOLOGY_NAME, "TridentTumblingCountWindowTest");
    JStormUnitTestRunner.submitTopology(tridentTopology.build(), null, 120, null);
}

Example 34 with Fields

use of backtype.storm.tuple.Fields in project jstorm by alibaba.

the class TridentWordCountTest method testTridentWordCount.

// to make sure the validator is right
@Test
public void testTridentWordCount() {
    LocalDRPC localDRPC = new LocalDRPC();
    FixedLimitBatchSpout spout = new FixedLimitBatchSpout(SPOUT_LIMIT, new Fields("sentence"), SPOUT_BATCH_SIZE, new Values("the cow jumped over the moon"), new Values("the man went to the store and bought some candy"), new Values("four score and seven years ago"), new Values("how many apples can you eat"), new Values("to be or not to be the person"));
    TridentTopology tridentTopology = new TridentTopology();
    TridentState wordCount = tridentTopology.newStream("spout", spout).parallelismHint(1).each(new Fields("sentence"), new Split(), new Fields("word")).groupBy(new Fields("word")).persistentAggregate(new MemoryMapState.Factory(), new Count(), new Fields("count")).parallelismHint(16);
    tridentTopology.newDRPCStream("words", localDRPC).each(new Fields("args"), new Split(), new Fields("keyword")).groupBy(new Fields("keyword")).stateQuery(wordCount, new Fields("keyword"), new MapGet(), new Fields("result")).each(new Fields("result"), new FilterNull()).aggregate(new Fields("result"), new Sum(), new Fields("sum"));
    Map config = new HashMap();
    config.put(Config.TOPOLOGY_NAME, "TridentWordCountTest");
    JStormUnitTestValidator validator = new JStormUnitTestDRPCValidator(localDRPC) {

        Logger LOG = LoggerFactory.getLogger(JStormUnitTestValidator.class);

        @Override
        public boolean validate(Map config) {
            String queryResult = executeLocalDRPC("words", "the");
            // the result is like [[8080]], so remove the [[]]
            queryResult = queryResult.substring(2, queryResult.length() - 2);
            // how many times of emit can finish a loop
            int oneLoopNeedEmits = (int) Math.ceil(SPOUT_CONTENT_TYPES / (float) SPOUT_BATCH_SIZE);
            // of all the spout content
            // the loop time of the LimitFixBatchSpout content
            int loopTime = SPOUT_LIMIT / oneLoopNeedEmits;
            int receiveCountOfThe = Integer.valueOf(queryResult);
            LOG.info("Final receive total " + receiveCountOfThe + " \"the\" when expected " + (loopTime * 5));
            // 5 "the" are in one loop
            boolean isCountOfTheRight = (receiveCountOfThe == loopTime * 5);
            // query the word count of these 3 words total
            queryResult = executeLocalDRPC("words", "be store kujou");
            queryResult = queryResult.substring(2, queryResult.length() - 2);
            int receiveCountOfBeAndStore = Integer.valueOf(queryResult);
            LOG.info("Final receive total " + receiveCountOfBeAndStore + " \"be\" and \"store\" and \"kujou\" " + "when expected " + (loopTime * 3));
            // 2 "be" 1 "store" 0 "kujou" are in one loop
            boolean isCountOfBeAndStoreRight = (receiveCountOfBeAndStore == loopTime * 3);
            return isCountOfTheRight && isCountOfBeAndStoreRight;
        }
    };
    try {
        boolean result = JStormUnitTestRunner.submitTopology(tridentTopology.build(), config, 90, validator);
        assertTrue("Topology should pass the validator", result);
    } finally {
        localDRPC.shutdown();
    }
}

Example 35 with Fields

use of backtype.storm.tuple.Fields in project jstorm by alibaba.

the class RollingTopWordsTest method testRollingTopWords.

@Test
public void testRollingTopWords() {
    TopologyBuilder topologyBuilder = new TopologyBuilder();
    topologyBuilder.setSpout("windowTestWordSpout", new WindowTestWordSpout(), 5);
    topologyBuilder.setBolt("windowTestRollingCountBolt", new WindowTestRollingCountBolt(9, 3), 4).fieldsGrouping("windowTestWordSpout", new Fields("word")).addConfiguration(Config.TOPOLOGY_TICK_TUPLE_FREQ_SECS, 3);
    topologyBuilder.setBolt("windowTestIntermediateRankingBolt", new WindowTestIntermediateRankingBolt(DEFAULT_COUNT), 4).fieldsGrouping("windowTestRollingCountBolt", new Fields("obj"));
    topologyBuilder.setBolt("windowTestTotalRankingsBolt", new WindowTestTotalRankingsBolt(DEFAULT_COUNT)).globalGrouping("windowTestIntermediateRankingBolt");
    Map config = new HashMap();
    config.put(Config.TOPOLOGY_NAME, "RollingTopWordsTest");
    // I really don't know how to validate if the result is right since
    // the tick time is not precise. It makes the output after passing
    // a window is unpredictable.
    // Now I just let it pass all the time.
    // TODO:FIX ME: how to validate if the result is right?
    JStormUnitTestRunner.submitTopology(topologyBuilder.createTopology(), config, 90, null);
}