Java JobConf.setReducerClass方法代码示例

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
/**
 * Run the test
 * 
 * @throws IOException on error
 */
public static void runTests() throws IOException {
  config.setLong("io.bytes.per.checksum", bytesPerChecksum);
  
  JobConf job = new JobConf(config, NNBench.class);

  job.setJobName("NNBench-" + operation);
  FileInputFormat.setInputPaths(job, new Path(baseDir, CONTROL_DIR_NAME));
  job.setInputFormat(SequenceFileInputFormat.class);
  
  // Explicitly set number of max map attempts to 1.
  job.setMaxMapAttempts(1);
  
  // Explicitly turn off speculative execution
  job.setSpeculativeExecution(false);

  job.setMapperClass(NNBenchMapper.class);
  job.setReducerClass(NNBenchReducer.class);

  FileOutputFormat.setOutputPath(job, new Path(baseDir, OUTPUT_DIR_NAME));
  job.setOutputKeyClass(Text.class);
  job.setOutputValueClass(Text.class);
  job.setNumReduceTasks((int) numberOfReduces);
  JobClient.runJob(job);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
private void runIOTest(
        Class<? extends Mapper<Text, LongWritable, Text, Text>> mapperClass, 
        Path outputDir) throws IOException {
  JobConf job = new JobConf(config, TestDFSIO.class);

  FileInputFormat.setInputPaths(job, getControlDir(config));
  job.setInputFormat(SequenceFileInputFormat.class);

  job.setMapperClass(mapperClass);
  job.setReducerClass(AccumulatingReducer.class);

  FileOutputFormat.setOutputPath(job, outputDir);
  job.setOutputKeyClass(Text.class);
  job.setOutputValueClass(Text.class);
  job.setNumReduceTasks(1);
  JobClient.runJob(job);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
/**
 * Sets up a job conf for the given job using the given config object. Ensures
 * that the correct input format is set, the mapper and and reducer class and
 * the input and output keys and value classes along with any other job
 * configuration.
 * 
 * @param config
 * @return JobConf representing the job to be ran
 * @throws IOException
 */
private JobConf getJob(ConfigExtractor config) throws IOException {
  JobConf job = new JobConf(config.getConfig(), SliveTest.class);
  job.setInputFormat(DummyInputFormat.class);
  FileOutputFormat.setOutputPath(job, config.getOutputPath());
  job.setMapperClass(SliveMapper.class);
  job.setPartitionerClass(SlivePartitioner.class);
  job.setReducerClass(SliveReducer.class);
  job.setOutputKeyClass(Text.class);
  job.setOutputValueClass(Text.class);
  job.setOutputFormat(TextOutputFormat.class);
  TextOutputFormat.setCompressOutput(job, false);
  job.setNumReduceTasks(config.getReducerAmount());
  job.setNumMapTasks(config.getMapAmount());
  return job;
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
private static void joinAs(String jointype,
    Class<? extends SimpleCheckerBase> c) throws Exception {
  final int srcs = 4;
  Configuration conf = new Configuration();
  JobConf job = new JobConf(conf, c);
  Path base = cluster.getFileSystem().makeQualified(new Path("/"+jointype));
  Path[] src = writeSimpleSrc(base, conf, srcs);
  job.set("mapreduce.join.expr", CompositeInputFormat.compose(jointype,
      SequenceFileInputFormat.class, src));
  job.setInt("testdatamerge.sources", srcs);
  job.setInputFormat(CompositeInputFormat.class);
  FileOutputFormat.setOutputPath(job, new Path(base, "out"));

  job.setMapperClass(c);
  job.setReducerClass(c);
  job.setOutputKeyClass(IntWritable.class);
  job.setOutputValueClass(IntWritable.class);
  JobClient.runJob(job);
  base.getFileSystem(job).delete(base, true);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
public void testEmptyJoin() throws Exception {
  JobConf job = new JobConf();
  Path base = cluster.getFileSystem().makeQualified(new Path("/empty"));
  Path[] src = { new Path(base,"i0"), new Path("i1"), new Path("i2") };
  job.set("mapreduce.join.expr", CompositeInputFormat.compose("outer",
      Fake_IF.class, src));
  job.setInputFormat(CompositeInputFormat.class);
  FileOutputFormat.setOutputPath(job, new Path(base, "out"));

  job.setMapperClass(IdentityMapper.class);
  job.setReducerClass(IdentityReducer.class);
  job.setOutputKeyClass(IncomparableKey.class);
  job.setOutputValueClass(NullWritable.class);

  JobClient.runJob(job);
  base.getFileSystem(job).delete(base, true);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
/**
 * Creates a simple copy job.
 * 
 * @param indirs List of input directories.
 * @param outdir Output directory.
 * @return JobConf initialised for a simple copy job.
 * @throws Exception If an error occurs creating job configuration.
 */
static JobConf createCopyJob(List<Path> indirs, Path outdir) throws Exception {

  Configuration defaults = new Configuration();
  JobConf theJob = new JobConf(defaults, TestJobControl.class);
  theJob.setJobName("DataMoveJob");

  FileInputFormat.setInputPaths(theJob, indirs.toArray(new Path[0]));
  theJob.setMapperClass(DataCopy.class);
  FileOutputFormat.setOutputPath(theJob, outdir);
  theJob.setOutputKeyClass(Text.class);
  theJob.setOutputValueClass(Text.class);
  theJob.setReducerClass(DataCopy.class);
  theJob.setNumMapTasks(12);
  theJob.setNumReduceTasks(4);
  return theJob;
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
/**
 * Use this before submitting a TableReduce job. It will
 * appropriately set up the JobConf.
 *
 * @param table  The output table.
 * @param reducer  The reducer class to use.
 * @param job  The current job configuration to adjust.
 * @param partitioner  Partitioner to use. Pass <code>null</code> to use
 * default partitioner.
 * @param addDependencyJars upload HBase jars and jars for any of the configured
 *           job classes via the distributed cache (tmpjars).
 * @throws IOException When determining the region count fails.
 */
public static void initTableReduceJob(String table,
  Class<? extends TableReduce> reducer, JobConf job, Class partitioner,
  boolean addDependencyJars) throws IOException {
  job.setOutputFormat(TableOutputFormat.class);
  job.setReducerClass(reducer);
  job.set(TableOutputFormat.OUTPUT_TABLE, table);
  job.setOutputKeyClass(ImmutableBytesWritable.class);
  job.setOutputValueClass(Put.class);
  job.setStrings("io.serializations", job.get("io.serializations"),
      MutationSerialization.class.getName(), ResultSerialization.class.getName());
  if (partitioner == HRegionPartitioner.class) {
    job.setPartitionerClass(HRegionPartitioner.class);
    int regions =
      MetaTableAccessor.getRegionCount(HBaseConfiguration.create(job), TableName.valueOf(table));
    if (job.getNumReduceTasks() > regions) {
      job.setNumReduceTasks(regions);
    }
  } else if (partitioner != null) {
    job.setPartitionerClass(partitioner);
  }
  if (addDependencyJars) {
    addDependencyJars(job);
  }
  initCredentials(job);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
@Override
protected void runJob(String jobName, Configuration c, List<Scan> scans)
    throws IOException, InterruptedException, ClassNotFoundException {
  JobConf job = new JobConf(TEST_UTIL.getConfiguration());

  job.setJobName(jobName);
  job.setMapperClass(Mapper.class);
  job.setReducerClass(Reducer.class);

  TableMapReduceUtil.initMultiTableSnapshotMapperJob(getSnapshotScanMapping(scans), Mapper.class,
      ImmutableBytesWritable.class, ImmutableBytesWritable.class, job, true, restoreDir);

  TableMapReduceUtil.addDependencyJars(job);

  job.setReducerClass(Reducer.class);
  job.setNumReduceTasks(1); // one to get final "first" and "last" key
  FileOutputFormat.setOutputPath(job, new Path(job.getJobName()));
  LOG.info("Started " + job.getJobName());

  RunningJob runningJob = JobClient.runJob(job);
  runningJob.waitForCompletion();
  assertTrue(runningJob.isSuccessful());
  LOG.info("After map/reduce completion - job " + jobName);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
public int run(String[] argv) throws IOException {
  if (argv.length < 2) {
    System.out.println("ExternalMapReduce <input> <output>");
    return -1;
  }
  Path outDir = new Path(argv[1]);
  Path input = new Path(argv[0]);
  JobConf testConf = new JobConf(getConf(), ExternalMapReduce.class);
  
  //try to load a class from libjar
  try {
    testConf.getClassByName("testjar.ClassWordCount");
  } catch (ClassNotFoundException e) {
    System.out.println("Could not find class from libjar");
    return -1;
  }
  
  
  testConf.setJobName("external job");
  FileInputFormat.setInputPaths(testConf, input);
  FileOutputFormat.setOutputPath(testConf, outDir);
  testConf.setMapperClass(MapClass.class);
  testConf.setReducerClass(Reduce.class);
  testConf.setNumReduceTasks(1);
  JobClient.runJob(testConf);
  return 0;
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
public static void runJobFail(JobConf conf, Path inDir, Path outDir)
       throws IOException, InterruptedException {
  conf.setJobName("test-job-fail");
  conf.setMapperClass(FailMapper.class);
  conf.setJarByClass(FailMapper.class);
  conf.setReducerClass(IdentityReducer.class);
  conf.setMaxMapAttempts(1);
  
  boolean success = runJob(conf, inDir, outDir, 1, 0);
  Assert.assertFalse("Job expected to fail succeeded", success);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
public static void runJobSucceed(JobConf conf, Path inDir, Path outDir)
       throws IOException, InterruptedException {
  conf.setJobName("test-job-succeed");
  conf.setMapperClass(IdentityMapper.class);
  //conf.setJar(new File(MiniMRYarnCluster.APPJAR).getAbsolutePath());
  conf.setReducerClass(IdentityReducer.class);
  
  boolean success = runJob(conf, inDir, outDir, 1 , 1);
  Assert.assertTrue("Job expected to succeed failed", success);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
@Test
public void testCombinerShouldUpdateTheReporter() throws Exception {
  JobConf conf = new JobConf(mrCluster.getConfig());
  int numMaps = 5;
  int numReds = 2;
  Path in = new Path(mrCluster.getTestWorkDir().getAbsolutePath(),
      "testCombinerShouldUpdateTheReporter-in");
  Path out = new Path(mrCluster.getTestWorkDir().getAbsolutePath(),
      "testCombinerShouldUpdateTheReporter-out");
  createInputOutPutFolder(in, out, numMaps);
  conf.setJobName("test-job-with-combiner");
  conf.setMapperClass(IdentityMapper.class);
  conf.setCombinerClass(MyCombinerToCheckReporter.class);
  //conf.setJarByClass(MyCombinerToCheckReporter.class);
  conf.setReducerClass(IdentityReducer.class);
  DistributedCache.addFileToClassPath(TestMRJobs.APP_JAR, conf);
  conf.setOutputCommitter(CustomOutputCommitter.class);
  conf.setInputFormat(TextInputFormat.class);
  conf.setOutputKeyClass(LongWritable.class);
  conf.setOutputValueClass(Text.class);

  FileInputFormat.setInputPaths(conf, in);
  FileOutputFormat.setOutputPath(conf, out);
  conf.setNumMapTasks(numMaps);
  conf.setNumReduceTasks(numReds);
  
  runJob(conf);
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
public static void doTestWithMapReduce(HBaseTestingUtility util, TableName tableName,
    String snapshotName, byte[] startRow, byte[] endRow, Path tableDir, int numRegions,
    int expectedNumSplits, boolean shutdownCluster) throws Exception {

  //create the table and snapshot
  createTableAndSnapshot(util, tableName, snapshotName, startRow, endRow, numRegions);

  if (shutdownCluster) {
    util.shutdownMiniHBaseCluster();
  }

  try {
    // create the job
    JobConf jobConf = new JobConf(util.getConfiguration());

    jobConf.setJarByClass(util.getClass());
    org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil.addDependencyJars(jobConf,
      TestTableSnapshotInputFormat.class);

    TableMapReduceUtil.initTableSnapshotMapJob(snapshotName, COLUMNS,
      TestTableSnapshotMapper.class, ImmutableBytesWritable.class,
      NullWritable.class, jobConf, true, tableDir);

    jobConf.setReducerClass(TestTableSnapshotInputFormat.TestTableSnapshotReducer.class);
    jobConf.setNumReduceTasks(1);
    jobConf.setOutputFormat(NullOutputFormat.class);

    RunningJob job = JobClient.runJob(jobConf);
    Assert.assertTrue(job.isSuccessful());
  } finally {
    if (!shutdownCluster) {
      util.getHBaseAdmin().deleteSnapshot(snapshotName);
      util.deleteTable(tableName);
    }
  }
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
/**
 * Based on args we submit the LoadGenerator as MR job.
 * Number of MapTasks is numMapTasks
 * @return exitCode for job submission
 */
private int submitAsMapReduce() {
  
  System.out.println("Running as a MapReduce job with " + 
      numMapTasks + " mapTasks;  Output to file " + mrOutDir);


  Configuration conf = new Configuration(getConf());
  
  // First set all the args of LoadGenerator as Conf vars to pass to MR tasks

  conf.set(LG_ROOT , root.toString());
  conf.setInt(LG_MAXDELAYBETWEENOPS, maxDelayBetweenOps);
  conf.setInt(LG_NUMOFTHREADS, numOfThreads);
  conf.set(LG_READPR, readProbs[0]+""); //Pass Double as string
  conf.set(LG_WRITEPR, writeProbs[0]+""); //Pass Double as string
  conf.setLong(LG_SEED, seed); //No idea what this is
  conf.setInt(LG_NUMMAPTASKS, numMapTasks);
  if (scriptFile == null && durations[0] <=0) {
    System.err.println("When run as a MapReduce job, elapsed Time or ScriptFile must be specified");
    System.exit(-1);
  }
  conf.setLong(LG_ELAPSEDTIME, durations[0]);
  conf.setLong(LG_STARTTIME, startTime); 
  if (scriptFile != null) {
    conf.set(LG_SCRIPTFILE , scriptFile);
  }
  conf.set(LG_FLAGFILE, flagFile.toString());
  
  // Now set the necessary conf variables that apply to run MR itself.
  JobConf jobConf = new JobConf(conf, LoadGenerator.class);
  jobConf.setJobName("NNLoadGeneratorViaMR");
  jobConf.setNumMapTasks(numMapTasks);
  jobConf.setNumReduceTasks(1); // 1 reducer to collect the results

  jobConf.setOutputKeyClass(Text.class);
  jobConf.setOutputValueClass(IntWritable.class);

  jobConf.setMapperClass(MapperThatRunsNNLoadGenerator.class);
  jobConf.setReducerClass(ReducerThatCollectsLGdata.class);

  jobConf.setInputFormat(DummyInputFormat.class);
  jobConf.setOutputFormat(TextOutputFormat.class);
  
  // Explicitly set number of max map attempts to 1.
  jobConf.setMaxMapAttempts(1);
  // Explicitly turn off speculative execution
  jobConf.setSpeculativeExecution(false);

  // This mapReduce job has no input but has output
  FileOutputFormat.setOutputPath(jobConf, new Path(mrOutDir));

  try {
    JobClient.runJob(jobConf);
  } catch (IOException e) {
    System.err.println("Failed to run job: " + e.getMessage());
    return -1;
  }
  return 0;
  
}
 

import org.apache.hadoop.mapred.JobConf; //导入方法依赖的package包/类
private static void setupPipesJob(JobConf conf) throws IOException {
  // default map output types to Text
  if (!getIsJavaMapper(conf)) {
    conf.setMapRunnerClass(PipesMapRunner.class);
    // Save the user's partitioner and hook in our's.
    setJavaPartitioner(conf, conf.getPartitionerClass());
    conf.setPartitionerClass(PipesPartitioner.class);
  }
  if (!getIsJavaReducer(conf)) {
    conf.setReducerClass(PipesReducer.class);
    if (!getIsJavaRecordWriter(conf)) {
      conf.setOutputFormat(NullOutputFormat.class);
    }
  }
  String textClassname = Text.class.getName();
  setIfUnset(conf, MRJobConfig.MAP_OUTPUT_KEY_CLASS, textClassname);
  setIfUnset(conf, MRJobConfig.MAP_OUTPUT_VALUE_CLASS, textClassname);
  setIfUnset(conf, MRJobConfig.OUTPUT_KEY_CLASS, textClassname);
  setIfUnset(conf, MRJobConfig.OUTPUT_VALUE_CLASS, textClassname);
  
  // Use PipesNonJavaInputFormat if necessary to handle progress reporting
  // from C++ RecordReaders ...
  if (!getIsJavaRecordReader(conf) && !getIsJavaMapper(conf)) {
    conf.setClass(Submitter.INPUT_FORMAT, 
                  conf.getInputFormat().getClass(), InputFormat.class);
    conf.setInputFormat(PipesNonJavaInputFormat.class);
  }
  
  String exec = getExecutable(conf);
  if (exec == null) {
    throw new IllegalArgumentException("No application program defined.");
  }
  // add default debug script only when executable is expressed as
  // <path>#<executable>
  if (exec.contains("#")) {
    // set default gdb commands for map and reduce task 
    String defScript = "$HADOOP_PREFIX/src/c++/pipes/debug/pipes-default-script";
    setIfUnset(conf, MRJobConfig.MAP_DEBUG_SCRIPT,defScript);
    setIfUnset(conf, MRJobConfig.REDUCE_DEBUG_SCRIPT,defScript);
  }
  URI[] fileCache = DistributedCache.getCacheFiles(conf);
  if (fileCache == null) {
    fileCache = new URI[1];
  } else {
    URI[] tmp = new URI[fileCache.length+1];
    System.arraycopy(fileCache, 0, tmp, 1, fileCache.length);
    fileCache = tmp;
  }
  try {
    fileCache[0] = new URI(exec);
  } catch (URISyntaxException e) {
    IOException ie = new IOException("Problem parsing execable URI " + exec);
    ie.initCause(e);
    throw ie;
  }
  DistributedCache.setCacheFiles(fileCache, conf);
}