Examples with MapReduceManager co.cask.cdap.test.MapReduceManager used on opensource projects

Example 1 with MapReduceManager

use of co.cask.cdap.test.MapReduceManager in project cdap by caskdata.

the class DataQualityAppTest method testMeanContentLength.

@Test
public void testMeanContentLength() throws Exception {
    Map<String, Set<String>> testMap = new HashMap<>();
    Set<String> testSet = new HashSet<>();
    testSet.add("Mean");
    testMap.put("content_length", testSet);
    DataQualityApp.DataQualityConfig config = new DataQualityApp.DataQualityConfig(WORKFLOW_SCHEDULE_MINUTES, getStreamSource(), "avg", testMap);
    ApplicationId appId = NamespaceId.DEFAULT.app("newApp2");
    AppRequest<DataQualityApp.DataQualityConfig> appRequest = new AppRequest<>(new ArtifactSummary(appArtifact.getArtifact(), appArtifact.getVersion()), config);
    ApplicationManager applicationManager = deployApplication(appId, appRequest);
    MapReduceManager mrManager = applicationManager.getMapReduceManager("FieldAggregator").start();
    mrManager.waitForRun(ProgramRunStatus.COMPLETED, 180, TimeUnit.SECONDS);
    ServiceManager serviceManager = applicationManager.getServiceManager(DataQualityService.SERVICE_NAME).start();
    serviceManager.waitForStatus(true);
    /* Test for aggregationsGetter handler */
    URL url = new URL(serviceManager.getServiceURL(), "v1/sources/logStream/fields/content_length/aggregations/Mean/timeseries");
    HttpResponse httpResponse = HttpRequests.execute(HttpRequest.get(url).build());
    Assert.assertEquals(HttpURLConnection.HTTP_OK, httpResponse.getResponseCode());
    String response = httpResponse.getResponseBodyAsString();
    List<TimestampValue> tsValueListActual = GSON.fromJson(response, TOKEN_TYPE_LIST_TIMESTAMP_VALUE);
    TimestampValue firstTimestampValue = tsValueListActual.get(0);
    Assert.assertEquals(256.0, firstTimestampValue.getValue());
    serviceManager.stop();
    serviceManager.waitForRun(ProgramRunStatus.KILLED, 180, TimeUnit.SECONDS);
}

Example 2 with MapReduceManager

use of co.cask.cdap.test.MapReduceManager in project cdap by caskdata.

the class DataQualityAppTest method testTotals.

@Test
public void testTotals() throws Exception {
    Map<String, Set<String>> testMap = new HashMap<>();
    Set<String> testSet = new HashSet<>();
    testSet.add("DiscreteValuesHistogram");
    testMap.put("content_length", testSet);
    testMap.put("status", testSet);
    testMap.put("request_time", testSet);
    DataQualityApp.DataQualityConfig config = new DataQualityApp.DataQualityConfig(WORKFLOW_SCHEDULE_MINUTES, getStreamSource(), "histogram", testMap);
    ApplicationId appId = NamespaceId.DEFAULT.app("newApp3");
    AppRequest<DataQualityApp.DataQualityConfig> appRequest = new AppRequest<>(new ArtifactSummary(appArtifact.getArtifact(), appArtifact.getVersion()), config);
    ApplicationManager applicationManager = deployApplication(appId, appRequest);
    MapReduceManager mrManager = applicationManager.getMapReduceManager("FieldAggregator").start();
    mrManager.waitForRun(ProgramRunStatus.COMPLETED, 180, TimeUnit.SECONDS);
    Map<String, Integer> expectedMap = new HashMap<>();
    expectedMap.put("256", 3);
    /* Test for the aggregationsGetter handler */
    ServiceManager serviceManager = applicationManager.getServiceManager(DataQualityService.SERVICE_NAME).start();
    serviceManager.waitForStatus(true);
    URL url = new URL(serviceManager.getServiceURL(), "v1/sources/logStream/fields/content_length/aggregations/DiscreteValuesHistogram/totals");
    HttpResponse httpResponse = HttpRequests.execute(HttpRequest.get(url).build());
    Assert.assertEquals(HttpURLConnection.HTTP_OK, httpResponse.getResponseCode());
    String response = httpResponse.getResponseBodyAsString();
    Map<String, Integer> histogramMap = GSON.fromJson(response, TOKEN_TYPE_MAP_STRING_INTEGER);
    Assert.assertEquals(expectedMap, histogramMap);
    /* Test for the fieldsGetter handler */
    url = new URL(serviceManager.getServiceURL(), "v1/sources/logStream/fields");
    httpResponse = HttpRequests.execute(HttpRequest.get(url).build());
    Assert.assertEquals(HttpURLConnection.HTTP_OK, httpResponse.getResponseCode());
    response = httpResponse.getResponseBodyAsString();
    Set<FieldDetail> outputSet = GSON.fromJson(response, TOKEN_TYPE_SET_FIELD_DETAIL);
    Set<FieldDetail> expectedSet = new HashSet<>();
    AggregationTypeValue aggregationTypeValue = new AggregationTypeValue("DiscreteValuesHistogram", true);
    Set<AggregationTypeValue> aggregationTypeValuesList = Sets.newHashSet(aggregationTypeValue);
    expectedSet.add(new FieldDetail("content_length", aggregationTypeValuesList));
    expectedSet.add(new FieldDetail("request_time", aggregationTypeValuesList));
    expectedSet.add(new FieldDetail("status", aggregationTypeValuesList));
    Assert.assertEquals(expectedSet, outputSet);
    /* Test for the aggregationTypesGetter handler */
    url = new URL(serviceManager.getServiceURL(), "v1/sources/logStream/fields/content_length");
    httpResponse = HttpRequests.execute(HttpRequest.get(url).build());
    Assert.assertEquals(HttpURLConnection.HTTP_OK, httpResponse.getResponseCode());
    response = httpResponse.getResponseBodyAsString();
    List<AggregationTypeValue> expectedAggregationTypeValuesList = new ArrayList<>();
    List<AggregationTypeValue> outputAggregationTypeValuesList = GSON.fromJson(response, TOKEN_TYPE_LIST_AGGREGATION_TYPE_VALUES);
    expectedAggregationTypeValuesList.add(new AggregationTypeValue("DiscreteValuesHistogram", true));
    Assert.assertEquals(expectedAggregationTypeValuesList, outputAggregationTypeValuesList);
    serviceManager.stop();
    serviceManager.waitForRun(ProgramRunStatus.KILLED, 180, TimeUnit.SECONDS);
}

Example 3 with MapReduceManager

use of co.cask.cdap.test.MapReduceManager in project cdap by caskdata.

the class DataCleansingMapReduceTest method testPartitionConsuming.

@Test
public void testPartitionConsuming() throws Exception {
    ApplicationManager applicationManager = deployApplication(DataCleansing.class);
    ServiceManager serviceManager = applicationManager.getServiceManager(DataCleansingService.NAME).start();
    serviceManager.waitForStatus(true);
    URL serviceURL = serviceManager.getServiceURL();
    // write a set of records to one partition and run the DataCleansingMapReduce job on that one partition
    createPartition(serviceURL, RECORD_SET1);
    // before starting the MR, there are 0 invalid records and 0 valid records, according to metrics
    Assert.assertEquals(0, getValidityMetrics(true));
    Assert.assertEquals(0, getValidityMetrics(false));
    Long now = System.currentTimeMillis();
    ImmutableMap<String, String> args = ImmutableMap.of(DataCleansingMapReduce.OUTPUT_PARTITION_KEY, now.toString(), DataCleansingMapReduce.SCHEMA_KEY, schemaJson);
    MapReduceManager mapReduceManager = applicationManager.getMapReduceManager(DataCleansingMapReduce.NAME).start(args);
    mapReduceManager.waitForRun(ProgramRunStatus.COMPLETED, 5, TimeUnit.MINUTES);
    compareData(now, DataCleansing.CLEAN_RECORDS, filterRecords(RECORD_SET1, true));
    compareData(now, DataCleansing.INVALID_RECORDS, filterRecords(RECORD_SET1, false));
    // assert that some of the records have indeed been filtered
    Assert.assertNotEquals(filterRecords(RECORD_SET1, true), RECORD_SET1);
    Assert.assertNotEquals(filterRecords(RECORD_SET1, false), Collections.<String>emptySet());
    // verify this via metrics
    Assert.assertEquals(1, getValidityMetrics(true));
    Assert.assertEquals(1, getValidityMetrics(false));
    // create two additional partitions
    createPartition(serviceURL, RECORD_SET2);
    createPartition(serviceURL, RECORD_SET3);
    // running the MapReduce job now processes these two new partitions (RECORD_SET1 and RECORD_SET2) and creates a new
    // partition with with the output
    now = System.currentTimeMillis();
    args = ImmutableMap.of(DataCleansingMapReduce.OUTPUT_PARTITION_KEY, now.toString(), DataCleansingMapReduce.SCHEMA_KEY, schemaJson);
    mapReduceManager = applicationManager.getMapReduceManager(DataCleansingMapReduce.NAME).start(args);
    mapReduceManager.waitForRuns(ProgramRunStatus.COMPLETED, 2, 5, TimeUnit.MINUTES);
    ImmutableSet<String> recordSets2and3 = ImmutableSet.<String>builder().addAll(RECORD_SET2).addAll(RECORD_SET3).build();
    compareData(now, DataCleansing.CLEAN_RECORDS, filterRecords(recordSets2and3, true));
    compareData(now, DataCleansing.INVALID_RECORDS, filterRecords(recordSets2and3, false));
    // verify this via metrics
    Assert.assertEquals(1, getValidityMetrics(true));
    Assert.assertEquals(5, getValidityMetrics(false));
    // running the MapReduce job without adding new partitions creates no additional output
    now = System.currentTimeMillis();
    args = ImmutableMap.of(DataCleansingMapReduce.OUTPUT_PARTITION_KEY, now.toString(), DataCleansingMapReduce.SCHEMA_KEY, schemaJson);
    mapReduceManager = applicationManager.getMapReduceManager(DataCleansingMapReduce.NAME).start(args);
    mapReduceManager.waitForRuns(ProgramRunStatus.COMPLETED, 3, 5, TimeUnit.MINUTES);
    compareData(now, DataCleansing.CLEAN_RECORDS, Collections.<String>emptySet());
    compareData(now, DataCleansing.INVALID_RECORDS, Collections.<String>emptySet());
    // verify that the records were properly partitioned on their zip
    DataSetManager<PartitionedFileSet> cleanRecords = getDataset(DataCleansing.CLEAN_RECORDS);
    PartitionFilter filter = PartitionFilter.builder().addValueCondition("zip", 84125).build();
    Assert.assertEquals(ImmutableSet.of(RECORD1, RECORD4, RECORD6), getDataFromFilter(cleanRecords.get(), filter));
    filter = PartitionFilter.builder().addValueCondition("zip", 84126).build();
    Assert.assertEquals(ImmutableSet.of(RECORD3, RECORD5), getDataFromFilter(cleanRecords.get(), filter));
}

Example 4 with MapReduceManager

use of co.cask.cdap.test.MapReduceManager in project cdap by caskdata.

the class LogAnalysisAppTest method test.

@Test
public void test() throws Exception {
    // Deploy the App
    ApplicationManager appManager = deployApplication(LogAnalysisApp.class);
    // Send a stream events to the Stream
    StreamManager streamManager = getStreamManager(LogAnalysisApp.LOG_STREAM);
    streamManager.send(LOG_1);
    streamManager.send(LOG_2);
    streamManager.send(LOG_3);
    // run the spark program
    SparkManager sparkManager = appManager.getSparkManager(LogAnalysisApp.ResponseCounterSpark.class.getSimpleName()).start();
    sparkManager.waitForRun(ProgramRunStatus.COMPLETED, 60, TimeUnit.SECONDS);
    // run the mapreduce job
    MapReduceManager mapReduceManager = appManager.getMapReduceManager(HitCounterProgram.class.getSimpleName()).start();
    mapReduceManager.waitForRun(ProgramRunStatus.COMPLETED, 3, TimeUnit.MINUTES);
    // start and wait for services
    ServiceManager hitCounterServiceManager = getServiceManager(appManager, LogAnalysisApp.HIT_COUNTER_SERVICE);
    ServiceManager responseCounterServiceManager = getServiceManager(appManager, LogAnalysisApp.RESPONSE_COUNTER_SERVICE);
    ServiceManager requestCounterServiceManager = getServiceManager(appManager, LogAnalysisApp.REQUEST_COUNTER_SERVICE);
    //Query for hit counts and verify it
    URL totalHitsURL = new URL(hitCounterServiceManager.getServiceURL(15, TimeUnit.SECONDS), LogAnalysisApp.HitCounterServiceHandler.HIT_COUNTER_SERVICE_PATH);
    HttpResponse response = HttpRequests.execute(HttpRequest.post(totalHitsURL).withBody("{\"url\":\"" + "/home.html" + "\"}").build());
    Assert.assertEquals(HttpURLConnection.HTTP_OK, response.getResponseCode());
    Assert.assertEquals(TOTAL_HITS_VALUE, response.getResponseBodyAsString());
    // query for total responses for a response code and verify it
    URL responseCodeURL = new URL(responseCounterServiceManager.getServiceURL(15, TimeUnit.SECONDS), LogAnalysisApp.ResponseCounterHandler.RESPONSE_COUNT_PATH + "/" + RESPONSE_CODE);
    HttpRequest request = HttpRequest.get(responseCodeURL).build();
    response = HttpRequests.execute(request);
    Assert.assertEquals(TOTAL_RESPONSE_VALUE, response.getResponseBodyAsString());
    // query to get partitions in the request count tpfs
    URL requestCountFilsetsURL = new URL(requestCounterServiceManager.getServiceURL(15, TimeUnit.SECONDS), LogAnalysisApp.RequestCounterHandler.REQUEST_COUNTER_PARTITIONS_PATH);
    request = HttpRequest.get(requestCountFilsetsURL).build();
    response = HttpRequests.execute(request);
    TreeSet<String> partitions = GSON.fromJson(response.getResponseBodyAsString(), new TypeToken<TreeSet<String>>() {
    }.getType());
    Assert.assertEquals(1, partitions.size());
    String partition = partitions.iterator().next();
    //Query for the contents of the files in this partition and verify
    URL requestFilesetContentURL = new URL(requestCounterServiceManager.getServiceURL(15, TimeUnit.SECONDS), LogAnalysisApp.RequestCounterHandler.REQUEST_FILE_CONTENT_PATH);
    response = HttpRequests.execute(HttpRequest.post(requestFilesetContentURL).withBody("{\"" + LogAnalysisApp.RequestCounterHandler.REQUEST_FILE_PATH_HANDLER_KEY + "\":\"" + partition + "\"}").build());
    Assert.assertEquals(HttpURLConnection.HTTP_OK, response.getResponseCode());
    Map<String, Integer> responseMap = GSON.fromJson(response.getResponseBodyAsString(), new TypeToken<Map<String, Integer>>() {
    }.getType());
    Assert.assertTrue(responseMap.equals(TPFS_RESULT));
}

Example 5 with MapReduceManager

use of co.cask.cdap.test.MapReduceManager in project cdap by caskdata.

the class SparkPageRankAppTest method test.

@Test
public void test() throws Exception {
    // Deploy the SparkPageRankApp
    ApplicationManager appManager = deployApplication(SparkPageRankApp.class);
    // Send a stream events to the Stream
    StreamManager streamManager = getStreamManager(SparkPageRankApp.BACKLINK_URL_STREAM);
    streamManager.send(Joiner.on(" ").join(URL_1, URL_2));
    streamManager.send(Joiner.on(" ").join(URL_1, URL_3));
    streamManager.send(Joiner.on(" ").join(URL_2, URL_1));
    streamManager.send(Joiner.on(" ").join(URL_3, URL_1));
    // Start service
    ServiceManager serviceManager = appManager.getServiceManager(SparkPageRankApp.SERVICE_HANDLERS).start();
    // Wait for service to start since the Spark program needs it
    serviceManager.waitForStatus(true);
    // Start the SparkPageRankProgram
    SparkManager sparkManager = appManager.getSparkManager(SparkPageRankApp.PageRankSpark.class.getSimpleName()).start();
    sparkManager.waitForFinish(60, TimeUnit.SECONDS);
    // Run RanksCounter which will count the number of pages for a pr
    MapReduceManager mapReduceManager = appManager.getMapReduceManager(SparkPageRankApp.RanksCounter.class.getSimpleName()).start();
    mapReduceManager.waitForFinish(3, TimeUnit.MINUTES);
    //Query for rank
    URL url = new URL(serviceManager.getServiceURL(15, TimeUnit.SECONDS), SparkPageRankApp.SparkPageRankServiceHandler.RANKS_PATH);
    HttpRequest request = HttpRequest.post(url).withBody(("{\"" + SparkPageRankApp.SparkPageRankServiceHandler.URL_KEY + "\":\"" + URL_1 + "\"}")).build();
    HttpResponse response = HttpRequests.execute(request);
    Assert.assertEquals(HttpURLConnection.HTTP_OK, response.getResponseCode());
    Assert.assertEquals(RANK, response.getResponseBodyAsString());
    // Request total pages for a page rank and verify it
    url = new URL(serviceManager.getServiceURL(15, TimeUnit.SECONDS), SparkPageRankApp.SparkPageRankServiceHandler.TOTAL_PAGES_PATH + "/" + RANK);
    response = HttpRequests.execute(HttpRequest.get(url).build());
    Assert.assertEquals(TOTAL_PAGES, response.getResponseBodyAsString());
}