Java Code Examples for org.apache.hadoop.mapred.JobConf#getCombinerClass()

The following examples show how to use org.apache.hadoop.mapred.JobConf#getCombinerClass() . 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: HadoopV2TaskContext.java From ignite with Apache License 2.0

5 votes

/**
 * @param taskInfo Task info.
 * @param job Job.
 * @param jobId Job ID.
 * @param locNodeId Local node ID.
 * @param jobConfDataInput DataInput for read JobConf.
 */
public HadoopV2TaskContext(HadoopTaskInfo taskInfo, HadoopJobEx job, HadoopJobId jobId,
    @Nullable UUID locNodeId, DataInput jobConfDataInput) throws IgniteCheckedException {
    super(taskInfo, job);
    this.locNodeId = locNodeId;

    // Before create JobConf instance we should set new context class loader.
    ClassLoader oldLdr = HadoopCommonUtils.setContextClassLoader(getClass().getClassLoader());

    try {
        JobConf jobConf = new JobConf();

        try {
            jobConf.readFields(jobConfDataInput);
        }
        catch (IOException e) {
            throw new IgniteCheckedException(e);
        }

        // For map-reduce jobs prefer local writes.
        jobConf.setBooleanIfUnset(PARAM_IGFS_PREFER_LOCAL_WRITES, true);

        initializePartiallyRawComparator(jobConf);

        jobCtx = new JobContextImpl(jobConf, new JobID(jobId.globalId().toString(), jobId.localId()));

        useNewMapper = jobConf.getUseNewMapper();
        useNewReducer = jobConf.getUseNewReducer();
        useNewCombiner = jobConf.getCombinerClass() == null;
    }
    finally {
        HadoopCommonUtils.restoreContextClassLoader(oldLdr);
    }
}

Example 2

Source File: TestShufflePlugin.java From hadoop with Apache License 2.0

4 votes

@Test
/**
 * A testing method verifying availability and accessibility of API that is needed
 * for sub-classes of ShuffleConsumerPlugin
 */
public void testConsumerApi() {

  JobConf jobConf = new JobConf();
  ShuffleConsumerPlugin<K, V> shuffleConsumerPlugin = new TestShuffleConsumerPlugin<K, V>();

  //mock creation
  ReduceTask mockReduceTask = mock(ReduceTask.class);
  TaskUmbilicalProtocol mockUmbilical = mock(TaskUmbilicalProtocol.class);
  Reporter mockReporter = mock(Reporter.class);
  FileSystem mockFileSystem = mock(FileSystem.class);
  Class<? extends org.apache.hadoop.mapred.Reducer>  combinerClass = jobConf.getCombinerClass();
  @SuppressWarnings("unchecked")  // needed for mock with generic
  CombineOutputCollector<K, V>  mockCombineOutputCollector =
    (CombineOutputCollector<K, V>) mock(CombineOutputCollector.class);
  org.apache.hadoop.mapreduce.TaskAttemptID mockTaskAttemptID =
    mock(org.apache.hadoop.mapreduce.TaskAttemptID.class);
  LocalDirAllocator mockLocalDirAllocator = mock(LocalDirAllocator.class);
  CompressionCodec mockCompressionCodec = mock(CompressionCodec.class);
  Counter mockCounter = mock(Counter.class);
  TaskStatus mockTaskStatus = mock(TaskStatus.class);
  Progress mockProgress = mock(Progress.class);
  MapOutputFile mockMapOutputFile = mock(MapOutputFile.class);
  Task mockTask = mock(Task.class);

  try {
    String [] dirs = jobConf.getLocalDirs();
    // verify that these APIs are available through super class handler
    ShuffleConsumerPlugin.Context<K, V> context =
   new ShuffleConsumerPlugin.Context<K, V>(mockTaskAttemptID, jobConf, mockFileSystem,
                                              mockUmbilical, mockLocalDirAllocator,
                                              mockReporter, mockCompressionCodec,
                                              combinerClass, mockCombineOutputCollector,
                                              mockCounter, mockCounter, mockCounter,
                                              mockCounter, mockCounter, mockCounter,
                                              mockTaskStatus, mockProgress, mockProgress,
                                              mockTask, mockMapOutputFile, null);
    shuffleConsumerPlugin.init(context);
    shuffleConsumerPlugin.run();
    shuffleConsumerPlugin.close();
  }
  catch (Exception e) {
    assertTrue("Threw exception:" + e, false);
  }

  // verify that these APIs are available for 3rd party plugins
  mockReduceTask.getTaskID();
  mockReduceTask.getJobID();
  mockReduceTask.getNumMaps();
  mockReduceTask.getPartition();
  mockReporter.progress();
}

Example 3

Source File: TestShuffleScheduler.java From hadoop with Apache License 2.0

4 votes

@SuppressWarnings("rawtypes")
@Test
public <K, V> void TestSucceedAndFailedCopyMap() throws Exception {
  JobConf job = new JobConf();
  job.setNumMapTasks(2);
  //mock creation
  TaskUmbilicalProtocol mockUmbilical = mock(TaskUmbilicalProtocol.class);
  Reporter mockReporter = mock(Reporter.class);
  FileSystem mockFileSystem = mock(FileSystem.class);
  Class<? extends org.apache.hadoop.mapred.Reducer>  combinerClass = job.getCombinerClass();
  @SuppressWarnings("unchecked")  // needed for mock with generic
  CombineOutputCollector<K, V>  mockCombineOutputCollector =
      (CombineOutputCollector<K, V>) mock(CombineOutputCollector.class);
  org.apache.hadoop.mapreduce.TaskAttemptID mockTaskAttemptID =
      mock(org.apache.hadoop.mapreduce.TaskAttemptID.class);
  LocalDirAllocator mockLocalDirAllocator = mock(LocalDirAllocator.class);
  CompressionCodec mockCompressionCodec = mock(CompressionCodec.class);
  Counter mockCounter = mock(Counter.class);
  TaskStatus mockTaskStatus = mock(TaskStatus.class);
  Progress mockProgress = mock(Progress.class);
  MapOutputFile mockMapOutputFile = mock(MapOutputFile.class);
  Task mockTask = mock(Task.class);
  @SuppressWarnings("unchecked")
  MapOutput<K, V> output = mock(MapOutput.class);

  ShuffleConsumerPlugin.Context<K, V> context =
      new ShuffleConsumerPlugin.Context<K, V>(
          mockTaskAttemptID, job, mockFileSystem,
          mockUmbilical, mockLocalDirAllocator,
          mockReporter, mockCompressionCodec,
          combinerClass, mockCombineOutputCollector,
          mockCounter, mockCounter, mockCounter,
          mockCounter, mockCounter, mockCounter,
          mockTaskStatus, mockProgress, mockProgress,
          mockTask, mockMapOutputFile, null);
  TaskStatus status = new TaskStatus() {
    @Override
    public boolean getIsMap() {
      return false;
    }
    @Override
    public void addFetchFailedMap(TaskAttemptID mapTaskId) {
    }
  };
  Progress progress = new Progress();
  ShuffleSchedulerImpl<K, V> scheduler = new ShuffleSchedulerImpl<K, V>(job,
      status, null, null, progress, context.getShuffledMapsCounter(),
      context.getReduceShuffleBytes(), context.getFailedShuffleCounter());

  MapHost host1 = new MapHost("host1", null);
  TaskAttemptID failedAttemptID = new TaskAttemptID(
      new org.apache.hadoop.mapred.TaskID(
      new JobID("test",0), TaskType.MAP, 0), 0);

  TaskAttemptID succeedAttemptID = new TaskAttemptID(
      new org.apache.hadoop.mapred.TaskID(
      new JobID("test",0), TaskType.MAP, 1), 1);

  // handle output fetch failure for failedAttemptID, part I
  scheduler.hostFailed(host1.getHostName());

  // handle output fetch succeed for succeedAttemptID
  long bytes = (long)500 * 1024 * 1024;
  scheduler.copySucceeded(succeedAttemptID, host1, bytes, 0, 500000, output);

  // handle output fetch failure for failedAttemptID, part II
  // for MAPREDUCE-6361: verify no NPE exception get thrown out
  scheduler.copyFailed(failedAttemptID, host1, true, false);
}

Example 4

Source File: TestShufflePlugin.java From big-c with Apache License 2.0

4 votes

@Test
/**
 * A testing method verifying availability and accessibility of API that is needed
 * for sub-classes of ShuffleConsumerPlugin
 */
public void testConsumerApi() {

  JobConf jobConf = new JobConf();
  ShuffleConsumerPlugin<K, V> shuffleConsumerPlugin = new TestShuffleConsumerPlugin<K, V>();

  //mock creation
  ReduceTask mockReduceTask = mock(ReduceTask.class);
  TaskUmbilicalProtocol mockUmbilical = mock(TaskUmbilicalProtocol.class);
  Reporter mockReporter = mock(Reporter.class);
  FileSystem mockFileSystem = mock(FileSystem.class);
  Class<? extends org.apache.hadoop.mapred.Reducer>  combinerClass = jobConf.getCombinerClass();
  @SuppressWarnings("unchecked")  // needed for mock with generic
  CombineOutputCollector<K, V>  mockCombineOutputCollector =
    (CombineOutputCollector<K, V>) mock(CombineOutputCollector.class);
  org.apache.hadoop.mapreduce.TaskAttemptID mockTaskAttemptID =
    mock(org.apache.hadoop.mapreduce.TaskAttemptID.class);
  LocalDirAllocator mockLocalDirAllocator = mock(LocalDirAllocator.class);
  CompressionCodec mockCompressionCodec = mock(CompressionCodec.class);
  Counter mockCounter = mock(Counter.class);
  TaskStatus mockTaskStatus = mock(TaskStatus.class);
  Progress mockProgress = mock(Progress.class);
  MapOutputFile mockMapOutputFile = mock(MapOutputFile.class);
  Task mockTask = mock(Task.class);

  try {
    String [] dirs = jobConf.getLocalDirs();
    // verify that these APIs are available through super class handler
    ShuffleConsumerPlugin.Context<K, V> context =
   new ShuffleConsumerPlugin.Context<K, V>(mockTaskAttemptID, jobConf, mockFileSystem,
                                              mockUmbilical, mockLocalDirAllocator,
                                              mockReporter, mockCompressionCodec,
                                              combinerClass, mockCombineOutputCollector,
                                              mockCounter, mockCounter, mockCounter,
                                              mockCounter, mockCounter, mockCounter,
                                              mockTaskStatus, mockProgress, mockProgress,
                                              mockTask, mockMapOutputFile, null);
    shuffleConsumerPlugin.init(context);
    shuffleConsumerPlugin.run();
    shuffleConsumerPlugin.close();
  }
  catch (Exception e) {
    assertTrue("Threw exception:" + e, false);
  }

  // verify that these APIs are available for 3rd party plugins
  mockReduceTask.getTaskID();
  mockReduceTask.getJobID();
  mockReduceTask.getNumMaps();
  mockReduceTask.getPartition();
  mockReporter.progress();
}

Example 5

Source File: TestShuffleScheduler.java From big-c with Apache License 2.0

4 votes

@SuppressWarnings("rawtypes")
@Test
public <K, V> void TestSucceedAndFailedCopyMap() throws Exception {
  JobConf job = new JobConf();
  job.setNumMapTasks(2);
  //mock creation
  TaskUmbilicalProtocol mockUmbilical = mock(TaskUmbilicalProtocol.class);
  Reporter mockReporter = mock(Reporter.class);
  FileSystem mockFileSystem = mock(FileSystem.class);
  Class<? extends org.apache.hadoop.mapred.Reducer>  combinerClass = job.getCombinerClass();
  @SuppressWarnings("unchecked")  // needed for mock with generic
  CombineOutputCollector<K, V>  mockCombineOutputCollector =
      (CombineOutputCollector<K, V>) mock(CombineOutputCollector.class);
  org.apache.hadoop.mapreduce.TaskAttemptID mockTaskAttemptID =
      mock(org.apache.hadoop.mapreduce.TaskAttemptID.class);
  LocalDirAllocator mockLocalDirAllocator = mock(LocalDirAllocator.class);
  CompressionCodec mockCompressionCodec = mock(CompressionCodec.class);
  Counter mockCounter = mock(Counter.class);
  TaskStatus mockTaskStatus = mock(TaskStatus.class);
  Progress mockProgress = mock(Progress.class);
  MapOutputFile mockMapOutputFile = mock(MapOutputFile.class);
  Task mockTask = mock(Task.class);
  @SuppressWarnings("unchecked")
  MapOutput<K, V> output = mock(MapOutput.class);

  ShuffleConsumerPlugin.Context<K, V> context =
      new ShuffleConsumerPlugin.Context<K, V>(
          mockTaskAttemptID, job, mockFileSystem,
          mockUmbilical, mockLocalDirAllocator,
          mockReporter, mockCompressionCodec,
          combinerClass, mockCombineOutputCollector,
          mockCounter, mockCounter, mockCounter,
          mockCounter, mockCounter, mockCounter,
          mockTaskStatus, mockProgress, mockProgress,
          mockTask, mockMapOutputFile, null);
  TaskStatus status = new TaskStatus() {
    @Override
    public boolean getIsMap() {
      return false;
    }
    @Override
    public void addFetchFailedMap(TaskAttemptID mapTaskId) {
    }
  };
  Progress progress = new Progress();
  ShuffleSchedulerImpl<K, V> scheduler = new ShuffleSchedulerImpl<K, V>(job,
      status, null, null, progress, context.getShuffledMapsCounter(),
      context.getReduceShuffleBytes(), context.getFailedShuffleCounter());

  MapHost host1 = new MapHost("host1", null);
  TaskAttemptID failedAttemptID = new TaskAttemptID(
      new org.apache.hadoop.mapred.TaskID(
      new JobID("test",0), TaskType.MAP, 0), 0);

  TaskAttemptID succeedAttemptID = new TaskAttemptID(
      new org.apache.hadoop.mapred.TaskID(
      new JobID("test",0), TaskType.MAP, 1), 1);

  // handle output fetch failure for failedAttemptID, part I
  scheduler.hostFailed(host1.getHostName());

  // handle output fetch succeed for succeedAttemptID
  long bytes = (long)500 * 1024 * 1024;
  scheduler.copySucceeded(succeedAttemptID, host1, bytes, 0, 500000, output);

  // handle output fetch failure for failedAttemptID, part II
  // for MAPREDUCE-6361: verify no NPE exception get thrown out
  scheduler.copyFailed(failedAttemptID, host1, true, false);
}