Java Code Examples for org.apache.hadoop.hbase.mapreduce.TableMapReduceUtil#addDependencyJarsForClasses()

private Job doVerify(Configuration conf, TableDescriptor tableDescriptor, String... auths)
    throws IOException, InterruptedException, ClassNotFoundException {
  Path outputDir = getTestDir(TEST_NAME, "verify-output");
  Job job = new Job(conf);
  job.setJarByClass(this.getClass());
  job.setJobName(TEST_NAME + " Verification for " + tableDescriptor.getTableName());
  setJobScannerConf(job);
  Scan scan = new Scan();
  scan.setAuthorizations(new Authorizations(auths));
  TableMapReduceUtil.initTableMapperJob(tableDescriptor.getTableName().getNameAsString(), scan,
      VerifyMapper.class, NullWritable.class, NullWritable.class, job);
  TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(), AbstractHBaseTool.class);
  int scannerCaching = conf.getInt("verify.scannercaching", SCANNER_CACHING);
  TableMapReduceUtil.setScannerCaching(job, scannerCaching);
  job.setNumReduceTasks(0);
  FileOutputFormat.setOutputPath(job, outputDir);
  assertTrue(job.waitForCompletion(true));
  return job;
}
 

protected Job doLoad(Configuration conf, TableDescriptor tableDescriptor) throws Exception {
  Path outputDir = getTestDir(TEST_NAME, "load-output");
  LOG.info("Load output dir: " + outputDir);

  NMapInputFormat.setNumMapTasks(conf, conf.getInt(NUM_MAP_TASKS_KEY, NUM_MAP_TASKS_DEFAULT));
  conf.set(TABLE_NAME_KEY, tableDescriptor.getTableName().getNameAsString());

  Job job = Job.getInstance(conf);
  job.setJobName(TEST_NAME + " Load for " + tableDescriptor.getTableName());
  job.setJarByClass(this.getClass());
  setMapperClass(job);
  job.setInputFormatClass(NMapInputFormat.class);
  job.setNumReduceTasks(0);
  setJobScannerConf(job);
  FileOutputFormat.setOutputPath(job, outputDir);

  TableMapReduceUtil.addDependencyJars(job);

  TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(), AbstractHBaseTool.class);
  TableMapReduceUtil.initCredentials(job);
  assertTrue(job.waitForCompletion(true));
  return job;
}
 

protected void doVerify(Configuration conf, TableDescriptor tableDescriptor) throws Exception {
  Path outputDir = getTestDir(TEST_NAME, "verify-output");
  LOG.info("Verify output dir: " + outputDir);

  Job job = Job.getInstance(conf);
  job.setJarByClass(this.getClass());
  job.setJobName(TEST_NAME + " Verification for " + tableDescriptor.getTableName());
  setJobScannerConf(job);

  Scan scan = new Scan();

  TableMapReduceUtil.initTableMapperJob(
      tableDescriptor.getTableName().getNameAsString(), scan, VerifyMapper.class,
      BytesWritable.class, BytesWritable.class, job);
  TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(), AbstractHBaseTool.class);
  int scannerCaching = conf.getInt("verify.scannercaching", SCANNER_CACHING);
  TableMapReduceUtil.setScannerCaching(job, scannerCaching);

  job.setReducerClass(VerifyReducer.class);
  job.setNumReduceTasks(conf.getInt(NUM_REDUCE_TASKS_KEY, NUM_REDUCE_TASKS_DEFAULT));
  FileOutputFormat.setOutputPath(job, outputDir);
  assertTrue(job.waitForCompletion(true));

  long numOutputRecords = job.getCounters().findCounter(Counters.ROWS_WRITTEN).getValue();
  assertEquals(0, numOutputRecords);
}
 

public int runRandomInputGenerator(int numMappers, long numNodes, Path tmpOutput,
    Integer width, Integer wrapMultiplier, Integer numWalkers)
    throws Exception {
  LOG.info("Running RandomInputGenerator with numMappers=" + numMappers
      + ", numNodes=" + numNodes);
  Job job = Job.getInstance(getConf());

  job.setJobName("Random Input Generator");
  job.setNumReduceTasks(0);
  job.setJarByClass(getClass());

  job.setInputFormatClass(GeneratorInputFormat.class);
  job.setOutputKeyClass(BytesWritable.class);
  job.setOutputValueClass(NullWritable.class);

  setJobConf(job, numMappers, numNodes, width, wrapMultiplier, numWalkers);

  job.setMapperClass(Mapper.class); //identity mapper

  FileOutputFormat.setOutputPath(job, tmpOutput);
  job.setOutputFormatClass(SequenceFileOutputFormat.class);
  TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(), Random64.class);

  boolean success = jobCompletion(job);

  return success ? 0 : 1;
}
 

public int runGenerator(int numMappers, long numNodes, Path tmpOutput,
    Integer width, Integer wrapMultiplier, Integer numWalkers)
    throws Exception {
  LOG.info("Running Generator with numMappers=" + numMappers +", numNodes=" + numNodes);
  createSchema();
  job = Job.getInstance(getConf());

  job.setJobName("Link Generator");
  job.setNumReduceTasks(0);
  job.setJarByClass(getClass());

  FileInputFormat.setInputPaths(job, tmpOutput);
  job.setInputFormatClass(OneFilePerMapperSFIF.class);
  job.setOutputKeyClass(NullWritable.class);
  job.setOutputValueClass(NullWritable.class);

  setJobConf(job, numMappers, numNodes, width, wrapMultiplier, numWalkers);

  setMapperForGenerator(job);

  job.setOutputFormatClass(NullOutputFormat.class);

  job.getConfiguration().setBoolean("mapreduce.map.speculative", false);
  TableMapReduceUtil.addDependencyJars(job);
  TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(),
                                                 AbstractHBaseTool.class);
  TableMapReduceUtil.initCredentials(job);

  boolean success = jobCompletion(job);

  return success ? 0 : 1;
}
 

private void configureJob() throws Exception {
    job = Job.getInstance(getConf(),
            "UpdateStatistics-" + tableName + "-" + snapshotName);
    PhoenixMapReduceUtil.setInput(job, NullDBWritable.class,
            snapshotName, tableName, restoreDir);

    PhoenixConfigurationUtil.setMRJobType(job.getConfiguration(), MRJobType.UPDATE_STATS);

    // DO NOT allow mapper splits using statistics since it may result into many smaller chunks
    PhoenixConfigurationUtil.setSplitByStats(job.getConfiguration(), false);

    job.setJarByClass(UpdateStatisticsTool.class);
    job.setMapperClass(TableSnapshotMapper.class);
    job.setMapOutputKeyClass(NullWritable.class);
    job.setMapOutputValueClass(NullWritable.class);
    job.setOutputFormatClass(NullOutputFormat.class);
    job.setNumReduceTasks(0);
    job.setPriority(this.jobPriority);

    TableMapReduceUtil.addDependencyJars(job);
    TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(), PhoenixConnection.class, Chronology.class,
            CharStream.class, TransactionSystemClient.class, TransactionNotInProgressException.class,
            ZKClient.class, DiscoveryServiceClient.class, ZKDiscoveryService.class,
            Cancellable.class, TTransportException.class, SpanReceiver.class, TransactionProcessor.class, Gauge.class, MetricRegistriesImpl.class);
    LOGGER.info("UpdateStatisticsTool running for: " + tableName
            + " on snapshot: " + snapshotName + " with restore dir: " + restoreDir);
}
 

public int run(Path outputDir, int numReducers) throws Exception {
  LOG.info("Running Verify with outputDir=" + outputDir +", numReducers=" + numReducers);

  job = Job.getInstance(getConf());

  job.setJobName("Link Verifier");
  job.setNumReduceTasks(numReducers);
  job.setJarByClass(getClass());

  setJobScannerConf(job);

  Scan scan = new Scan();
  scan.addColumn(FAMILY_NAME, COLUMN_PREV);
  scan.setCaching(10000);
  scan.setCacheBlocks(false);
  if (isMultiUnevenColumnFamilies(getConf())) {
    scan.addColumn(BIG_FAMILY_NAME, BIG_FAMILY_NAME);
    scan.addColumn(TINY_FAMILY_NAME, TINY_FAMILY_NAME);
  }

  TableMapReduceUtil.initTableMapperJob(getTableName(getConf()).getName(), scan,
      VerifyMapper.class, BytesWritable.class, BytesWritable.class, job);
  TableMapReduceUtil.addDependencyJarsForClasses(job.getConfiguration(),
                                                 AbstractHBaseTool.class);

  job.getConfiguration().setBoolean("mapreduce.map.speculative", false);

  job.setReducerClass(VerifyReducer.class);
  job.setOutputFormatClass(SequenceFileAsBinaryOutputFormat.class);
  job.setOutputKeyClass(BytesWritable.class);
  job.setOutputValueClass(BytesWritable.class);
  TextOutputFormat.setOutputPath(job, outputDir);

  boolean success = job.waitForCompletion(true);

  if (success) {
    Counters counters = job.getCounters();
    if (null == counters) {
      LOG.warn("Counters were null, cannot verify Job completion."
          + " This is commonly a result of insufficient YARN configuration.");
      // We don't have access to the counters to know if we have "bad" counts
      return 0;
    }

    // If we find no unexpected values, the job didn't outright fail
    if (verifyUnexpectedValues(counters)) {
      // We didn't check referenced+unreferenced counts, leave that to visual inspection
      return 0;
    }
  }

  // We failed
  return 1;
}