Example usage for org.apache.hadoop.mapred JobConf setNumMapTasks

List of usage examples for org.apache.hadoop.mapred JobConf setNumMapTasks

  1. HOME
  2. Java
  3. org.apache.hadoop
  4. org.apache.hadoop.mapred.*
  5. JobConf
  6. setNumMapTasks

Introduction

In this page you can find the example usage for org.apache.hadoop.mapred JobConf setNumMapTasks.

Prototype

public void setNumMapTasks(int n)

Source Link

Document

Set the number of map tasks for this job.

Usage

From source file:edu.umd.cloud9.collection.clue.CountClueWarcRecords.java

License:Apache License

/**
 * Runs this tool.//from   w  w w.j a  va2  s .com
 */
@SuppressWarnings("static-access")
public int run(String[] args) throws Exception {
    Options options = new Options();
    options.addOption(new Option(ORIGINAL_OPTION, "use original ClueWeb09 distribution"));
    options.addOption(new Option(REPACKED_OPTION, "use repacked SequenceFiles"));

    options.addOption(OptionBuilder.withArgName("path").hasArg()
            .withDescription("path: base path for 'original', actual path for 'repacked'").create(PATH_OPTION));
    options.addOption(OptionBuilder.withArgName("path").hasArg().withDescription("DocnoMapping data path")
            .create(MAPPING_OPTION));
    options.addOption(OptionBuilder.withArgName("num").hasArg()
            .withDescription("segment number (required if 'original')").create(SEGMENT_OPTION));
    options.addOption(OptionBuilder.withArgName("path").hasArg()
            .withDescription("output file to write the number of records").create(COUNT_OPTION));

    CommandLine cmdline;
    CommandLineParser parser = new GnuParser();
    try {
        cmdline = parser.parse(options, args);
    } catch (ParseException exp) {
        HelpFormatter formatter = new HelpFormatter();
        formatter.printHelp(this.getClass().getName(), options);
        ToolRunner.printGenericCommandUsage(System.out);
        System.err.println("Error parsing command line: " + exp.getMessage());
        return -1;
    }

    boolean repacked;
    if (cmdline.hasOption(REPACKED_OPTION)) {
        repacked = true;
    } else if (cmdline.hasOption(ORIGINAL_OPTION)) {
        repacked = false;
    } else {
        HelpFormatter formatter = new HelpFormatter();
        formatter.printHelp(this.getClass().getName(), options);
        ToolRunner.printGenericCommandUsage(System.out);
        System.err.println("Expecting either -original or -repacked");
        return -1;
    }

    if (!cmdline.hasOption(PATH_OPTION) || !cmdline.hasOption(MAPPING_OPTION)
            || (!repacked && !cmdline.hasOption(SEGMENT_OPTION))) {
        HelpFormatter formatter = new HelpFormatter();
        formatter.printHelp(this.getClass().getName(), options);
        ToolRunner.printGenericCommandUsage(System.out);
        return -1;
    }

    String path = cmdline.getOptionValue(PATH_OPTION);
    String mappingFile = cmdline.getOptionValue(MAPPING_OPTION);

    int segment = 1;
    if (!repacked) {
        segment = Integer.parseInt(cmdline.getOptionValue(SEGMENT_OPTION));
    }

    LOG.info("Tool name: " + CountClueWarcRecords.class.getSimpleName());
    LOG.info(" - repacked: " + repacked);
    LOG.info(" - path: " + path);
    LOG.info(" - mapping file: " + mappingFile);
    if (!repacked) {
        LOG.info(" - segment number: " + segment);
    }

    FileSystem fs = FileSystem.get(getConf());
    int mapTasks = 10;

    JobConf conf = new JobConf(getConf(), CountClueWarcRecords.class);
    conf.setJobName(
            CountClueWarcRecords.class.getSimpleName() + (repacked ? ":" + path : ":segment" + segment));

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(0);

    if (repacked) {
        // Note, we have to add the files one by one, otherwise, SequenceFileInputFormat
        // thinks its a MapFile.
        for (FileStatus status : fs.listStatus(new Path(path))) {
            FileInputFormat.addInputPath(conf, status.getPath());
        }
    } else {
        ClueCollectionPathConstants.addEnglishCollectionPart(conf, path, segment);
    }

    DistributedCache.addCacheFile(new URI(mappingFile), conf);

    if (repacked) {
        conf.setInputFormat(SequenceFileInputFormat.class);
    } else {
        conf.setInputFormat(ClueWarcInputFormat.class);
    }

    conf.setOutputFormat(NullOutputFormat.class);
    conf.setMapperClass(MyMapper.class);

    RunningJob job = JobClient.runJob(conf);
    Counters counters = job.getCounters();
    int numDocs = (int) counters.findCounter(Records.PAGES).getCounter();

    LOG.info("Read " + numDocs + " docs.");

    if (cmdline.hasOption(COUNT_OPTION)) {
        String f = cmdline.getOptionValue(COUNT_OPTION);
        FSDataOutputStream out = fs.create(new Path(f));
        out.write(new Integer(numDocs).toString().getBytes());
        out.close();
    }

    return 0;
}

From source file:edu.umd.cloud9.collection.clue.RepackClueWarcRecords.java

License:Apache License

/**
 * Runs this tool./*from www  .j  a v a 2s.com*/
 */
public int run(String[] args) throws Exception {
    if (args.length != 5) {
        printUsage();
        return -1;
    }

    String basePath = args[0];
    String outputPath = args[1];
    int segment = Integer.parseInt(args[2]);
    String data = args[3];
    String compressionType = args[4];

    if (!compressionType.equals("block") && !compressionType.equals("record")
            && !compressionType.equals("none")) {
        System.err.println("Error: \"" + compressionType + "\" unknown compression type!");
        System.exit(-1);
    }

    // Default block size.
    int blocksize = 1000000;

    JobConf conf = new JobConf(RepackClueWarcRecords.class);
    conf.setJobName("RepackClueWarcRecords:segment" + segment);

    conf.set("DocnoMappingDataFile", data);

    LOG.info("Tool name: RepackClueWarcRecords");
    LOG.info(" - base path: " + basePath);
    LOG.info(" - output path: " + outputPath);
    LOG.info(" - segment number: " + segment);
    LOG.info(" - docno mapping data file: " + data);
    LOG.info(" - compression type: " + compressionType);

    if (compressionType.equals("block")) {
        LOG.info(" - block size: " + blocksize);
    }

    int mapTasks = 10;

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(0);

    ClueCollectionPathConstants.addEnglishCollectionPart(conf, basePath, segment);

    SequenceFileOutputFormat.setOutputPath(conf, new Path(outputPath));

    if (compressionType.equals("none")) {
        SequenceFileOutputFormat.setCompressOutput(conf, false);
    } else {
        SequenceFileOutputFormat.setCompressOutput(conf, true);

        if (compressionType.equals("record")) {
            SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.RECORD);
        } else {
            SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);
            conf.setInt("io.seqfile.compress.blocksize", blocksize);
        }
    }

    conf.setInputFormat(ClueWarcInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);
    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(ClueWarcRecord.class);

    conf.setMapperClass(MyMapper.class);

    // Delete the output directory if it exists already.
    FileSystem.get(conf).delete(new Path(outputPath), true);

    JobClient.runJob(conf);

    return 0;
}

From source file:edu.umd.cloud9.collection.line.NumberTextDocuments.java

License:Apache License

/**
 * Runs this tool.//from   ww w .j a v  a 2  s .c  o m
 */
public int run(String[] args) throws Exception {
    if (args.length != 4) {
        printUsage();
        return -1;
    }

    String inputPath = args[0];
    String outputPath = args[1];
    String outputFile = args[2];
    int mapTasks = Integer.parseInt(args[3]);

    sLogger.info("Tool: NumberTextDocuments");
    sLogger.info(" - Input path: " + inputPath);
    sLogger.info(" - Output path: " + outputPath);
    sLogger.info(" - Output file: " + outputFile);
    sLogger.info("Launching with " + mapTasks + " mappers...");

    JobConf conf = new JobConf(getConf(), NumberTextDocuments.class);
    conf.setJobName("NumberTextDocuments");

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(1);

    FileInputFormat.setInputPaths(conf, new Path(inputPath));
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));
    FileOutputFormat.setCompressOutput(conf, false);

    conf.setInputFormat(TextDocumentInputFormat.class);
    conf.setOutputKeyClass(Text.class);
    conf.setOutputValueClass(IntWritable.class);
    conf.setOutputFormat(TextOutputFormat.class);

    conf.setMapperClass(MyMapper.class);
    conf.setReducerClass(MyReducer.class);

    // delete the output directory if it exists already
    FileSystem.get(conf).delete(new Path(outputPath), true);

    JobClient.runJob(conf);

    String input = outputPath + (outputPath.endsWith("/") ? "" : "/") + "/part-00000";
    TextDocnoMapping.writeDocnoData(input, outputFile, FileSystem.get(getConf()));

    return 0;
}

From source file:edu.umd.cloud9.collection.medline.NumberMedlineCitations.java

License:Apache License

/**
 * Runs this tool./*from   w  ww .  j a  v a2 s. c  o  m*/
 */
public int run(String[] args) throws Exception {
    if (args.length != 4) {
        printUsage();
        return -1;
    }

    String inputPath = args[0];
    String outputPath = args[1];
    String outputFile = args[2];
    int mapTasks = Integer.parseInt(args[3]);

    sLogger.info("Tool name: NumberMedlineCitations");
    sLogger.info(" - Input path: " + inputPath);
    sLogger.info(" - Output path: " + outputPath);
    sLogger.info(" - Output file: " + outputFile);
    sLogger.info("Launching with " + mapTasks + " mappers...");

    JobConf conf = new JobConf(getConf(), NumberMedlineCitations.class);
    conf.setJobName("NumberMedlineCitations");

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(1);

    FileInputFormat.setInputPaths(conf, new Path(inputPath));
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));
    FileOutputFormat.setCompressOutput(conf, false);

    conf.setInputFormat(MedlineCitationInputFormat.class);
    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(IntWritable.class);
    conf.setOutputFormat(TextOutputFormat.class);

    conf.setMapperClass(MyMapper.class);
    conf.setReducerClass(MyReducer.class);

    // delete the output directory if it exists already
    FileSystem.get(conf).delete(new Path(outputPath), true);

    RunningJob job = JobClient.runJob(conf);

    // write out various properties
    Counters counters = job.getCounters();
    Counter counter = counters.findCounter("edu.umd.cloud9.collection.medline.NumberMedlineCitations$Citations",
            0, "");

    int numdocs = (int) counter.getCounter();
    sLogger.info("total number of docs: " + numdocs);

    MedlineDocnoMapping.writeDocidData(outputPath + "/part-00000", outputFile);

    return 0;
}

From source file:edu.umd.cloud9.collection.trec.NumberTrecDocuments.java

License:Apache License

/**
 * Runs this tool./*from w  ww .  j  a v a  2s . c  om*/
 */
public int run(String[] args) throws Exception {
    if (args.length != 3) {
        printUsage();
        return -1;
    }

    String inputPath = args[0];
    String outputPath = args[1];
    String outputFile = args[2];

    LOG.info("Tool: " + NumberTrecDocuments.class.getCanonicalName());
    LOG.info(" - Input path: " + inputPath);
    LOG.info(" - Output path: " + outputPath);
    LOG.info(" - Output file: " + outputFile);

    JobConf conf = new JobConf(getConf(), NumberTrecDocuments.class);
    conf.setJobName(NumberTrecDocuments.class.getSimpleName());

    conf.setNumMapTasks(100); // Arbitrary number; doesn't matter.
    conf.setNumReduceTasks(1);

    FileInputFormat.setInputPaths(conf, new Path(inputPath));
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));
    FileOutputFormat.setCompressOutput(conf, false);

    conf.setInputFormat(TrecDocumentInputFormat.class);
    conf.setOutputKeyClass(Text.class);
    conf.setOutputValueClass(IntWritable.class);
    conf.setOutputFormat(TextOutputFormat.class);

    conf.setMapperClass(MyMapper.class);
    conf.setReducerClass(MyReducer.class);

    // delete the output directory if it exists already
    FileSystem.get(conf).delete(new Path(outputPath), true);

    JobClient.runJob(conf);

    String input = outputPath + (outputPath.endsWith("/") ? "" : "/") + "/part-00000";
    TrecDocnoMapping.writeMappingData(new Path(input), new Path(outputFile), FileSystem.get(getConf()));

    return 0;
}

From source file:edu.umd.cloud9.collection.trecweb.NumberTrecWebDocuments.java

License:Apache License

public int run(String[] args) throws Exception {
    if (args.length != 4) {
        System.out.println("usage: [input] [output-dir] [output-file] [num-mappers]");
        System.exit(-1);// w ww.j  a v a  2  s  . co  m
    }

    String inputPath = args[0];
    String outputPath = args[1];
    String outputFile = args[2];
    int mapTasks = Integer.parseInt(args[3]);

    LOG.info("Tool name: " + NumberTrecWebDocuments.class.getCanonicalName());
    LOG.info(" - input path: " + inputPath);
    LOG.info(" - output path: " + outputPath);
    LOG.info(" - output file: " + outputFile);
    LOG.info(" - number of mappers: " + mapTasks);

    JobConf conf = new JobConf(getConf(), NumberTrecWebDocuments.class);
    conf.setJobName(NumberTrecWebDocuments.class.getSimpleName());

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(1);

    FileInputFormat.setInputPaths(conf, new Path(inputPath));
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));
    FileOutputFormat.setCompressOutput(conf, false);

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputKeyClass(Text.class);
    conf.setOutputValueClass(IntWritable.class);
    conf.setOutputFormat(TextOutputFormat.class);

    conf.setMapperClass(MyMapper.class);
    conf.setReducerClass(MyReducer.class);

    // Delete the output directory if it exists already.
    FileSystem.get(conf).delete(new Path(outputPath), true);

    JobClient.runJob(conf);

    writeMappingData(new Path(outputPath + "/part-00000"), new Path(outputFile), FileSystem.get(conf));

    return 0;
}

From source file:edu.umd.cloud9.collection.trecweb.RepackGov2Documents.java

License:Apache License

/**
 * Runs this tool.//from w ww  .  j  a va  2s .  c o  m
 */
public int run(String[] args) throws Exception {
    if (args.length != 3) {
        printUsage();
        return -1;
    }

    String basePath = args[0];
    String outputPath = args[1];
    String compressionType = args[2];

    if (!compressionType.equals("block") && !compressionType.equals("record")
            && !compressionType.equals("none")) {
        System.err.println("Error: \"" + compressionType + "\" unknown compression type!");
        System.exit(-1);
    }

    // this is the default block size
    int blocksize = 1000000;

    JobConf conf = new JobConf(RepackGov2Documents.class);
    conf.setJobName("RepackGov2Documents");

    sLogger.info("Tool name: RepackGov2Documents");
    sLogger.info(" - base path: " + basePath);
    sLogger.info(" - output path: " + outputPath);
    sLogger.info(" - compression type: " + compressionType);

    if (compressionType.equals("block")) {
        sLogger.info(" - block size: " + blocksize);
    }

    int mapTasks = 10;

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(500);

    // 272
    for (int i = 0; i <= 272; i++) {
        String path = basePath + "/GX";
        String indexNum = Integer.toString(i);
        if (indexNum.length() == 1) {
            path += "00";
        }
        if (indexNum.length() == 2) {
            path += "0";
        }
        path += indexNum;
        FileInputFormat.addInputPath(conf, new Path(path));
    }

    SequenceFileOutputFormat.setOutputPath(conf, new Path(outputPath));

    if (compressionType.equals("none")) {
        SequenceFileOutputFormat.setCompressOutput(conf, false);
    } else {
        SequenceFileOutputFormat.setCompressOutput(conf, true);

        if (compressionType.equals("record")) {
            SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.RECORD);
        } else {
            SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);
            conf.setInt("io.seqfile.compress.blocksize", blocksize);
        }
    }

    conf.setInputFormat(TrecWebDocumentInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);
    conf.setOutputKeyClass(LongWritable.class);
    conf.setOutputValueClass(TrecWebDocument.class);

    conf.setMapperClass(MyMapper.class);

    // delete the output directory if it exists already
    FileSystem.get(conf).delete(new Path(outputPath), true);

    JobClient.runJob(conf);

    return 0;
}

From source file:edu.umd.cloud9.collection.trecweb.RepackWt10gDocuments.java

License:Apache License

/**
 * Runs this tool.//from  w  w w  .  j a  va  2  s . com
 */
public int run(String[] args) throws Exception {
    if (args.length != 3) {
        printUsage();
        return -1;
    }

    String basePath = args[0];
    String outputPath = args[1];
    String compressionType = args[2];

    if (!compressionType.equals("block") && !compressionType.equals("record")
            && !compressionType.equals("none")) {
        System.err.println("Error: \"" + compressionType + "\" unknown compression type!");
        System.exit(-1);
    }

    // this is the default block size
    int blocksize = 1000000;

    JobConf conf = new JobConf(RepackWt10gDocuments.class);
    conf.setJobName("RepackWt10gDocuments");

    sLogger.info("Tool name: RepackWt10gDocuments");
    sLogger.info(" - base path: " + basePath);
    sLogger.info(" - output path: " + outputPath);
    sLogger.info(" - compression type: " + compressionType);

    if (compressionType.equals("block")) {
        sLogger.info(" - block size: " + blocksize);
    }

    int mapTasks = 10;

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(50);

    for (int i = 1; i <= 104; i++) {
        String path = basePath + "/WTX";
        String indexNum = Integer.toString(i);
        if (indexNum.length() == 1) {
            path += "00";
        }
        if (indexNum.length() == 2) {
            path += "0";
        }
        path += indexNum;
        FileInputFormat.addInputPath(conf, new Path(path));
    }

    SequenceFileOutputFormat.setOutputPath(conf, new Path(outputPath));

    if (compressionType.equals("none")) {
        SequenceFileOutputFormat.setCompressOutput(conf, false);
    } else {
        SequenceFileOutputFormat.setCompressOutput(conf, true);

        if (compressionType.equals("record")) {
            SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.RECORD);
        } else {
            SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);
            conf.setInt("io.seqfile.compress.blocksize", blocksize);
        }
    }

    conf.setInputFormat(TrecWebDocumentInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);
    conf.setOutputKeyClass(LongWritable.class);
    conf.setOutputValueClass(TrecWebDocument.class);

    conf.setMapperClass(MyMapper.class);

    // delete the output directory if it exists already
    FileSystem.get(conf).delete(new Path(outputPath), true);

    JobClient.runJob(conf);

    return 0;
}

From source file:edu.umd.cloud9.collection.wikipedia.DemoCountWikipediaPages.java

License:Apache License

@SuppressWarnings("static-access")
@Override//from  ww  w  .j a  v  a  2  s .  com
public int run(String[] args) throws Exception {
    Options options = new Options();
    options.addOption(
            OptionBuilder.withArgName("path").hasArg().withDescription("XML dump file").create(INPUT_OPTION));

    CommandLine cmdline;
    CommandLineParser parser = new GnuParser();
    try {
        cmdline = parser.parse(options, args);
    } catch (ParseException exp) {
        System.err.println("Error parsing command line: " + exp.getMessage());
        return -1;
    }

    if (!cmdline.hasOption(INPUT_OPTION)) {
        HelpFormatter formatter = new HelpFormatter();
        formatter.printHelp(this.getClass().getName(), options);
        ToolRunner.printGenericCommandUsage(System.out);
        return -1;
    }

    String inputPath = cmdline.getOptionValue(INPUT_OPTION);

    LOG.info("Tool name: " + this.getClass().getName());
    LOG.info(" - XML dump file: " + inputPath);

    JobConf conf = new JobConf(getConf(), DemoCountWikipediaPages.class);
    conf.setJobName(String.format("DemoCountWikipediaPages[%s: %s]", INPUT_OPTION, inputPath));

    conf.setNumMapTasks(10);
    conf.setNumReduceTasks(0);

    FileInputFormat.setInputPaths(conf, new Path(inputPath));

    conf.setInputFormat(WikipediaPageInputFormat.class);
    conf.setOutputFormat(NullOutputFormat.class);

    conf.setMapperClass(MyMapper.class);

    JobClient.runJob(conf);

    return 0;
}

From source file:edu.umd.cloud9.demo.DemoWordCondProbJSON.java

License:Apache License

/**
 * Runs this tool./*from   ww  w .  ja  v a 2 s .  co  m*/
 */
public int run(String[] args) throws Exception {
    if (args.length != 4) {
        printUsage();
        return -1;
    }

    String inputPath = args[0];
    String outputPath = args[1];

    int mapTasks = Integer.parseInt(args[2]);
    int reduceTasks = Integer.parseInt(args[3]);

    sLogger.info("Tool: DemoWordCondProbJSON");
    sLogger.info(" - input path: " + inputPath);
    sLogger.info(" - output path: " + outputPath);
    sLogger.info(" - number of mappers: " + mapTasks);
    sLogger.info(" - number of reducers: " + reduceTasks);

    JobConf conf = new JobConf(DemoWordCondProbJSON.class);
    conf.setJobName("DemoWordCondProbJSON");

    conf.setNumMapTasks(mapTasks);
    conf.setNumReduceTasks(reduceTasks);

    FileInputFormat.setInputPaths(conf, new Path(inputPath));
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));
    FileOutputFormat.setCompressOutput(conf, false);

    conf.setOutputKeyClass(MyTuple.class);
    conf.setOutputValueClass(FloatWritable.class);
    conf.setOutputFormat(TextOutputFormat.class);

    conf.setMapperClass(MyMapper.class);
    // this is a potential gotcha! can't use ReduceClass for combine because
    // we have not collected all the counts yet, so we can't divide through
    // to compute the conditional probabilities
    conf.setCombinerClass(IdentityReducer.class);
    conf.setReducerClass(MyReducer.class);
    conf.setPartitionerClass(MyPartitioner.class);

    // Delete the output directory if it exists already
    Path outputDir = new Path(outputPath);
    FileSystem.get(conf).delete(outputDir, true);

    long startTime = System.currentTimeMillis();
    JobClient.runJob(conf);
    sLogger.info("Job Finished in " + (System.currentTimeMillis() - startTime) / 1000.0 + " seconds");

    return 0;
}