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

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

Introduction

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

Prototype

public void setMapperClass(Class<? extends Mapper> theClass) 

Source Link

Document

Set the Mapper class for the job.

Usage

From source file:edu.umd.cloud9.pagerank.FindMaxPageRankNodes.java

License:Apache License

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

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

    sLogger.info("Tool name: FindMaxPageRankNodes");
    sLogger.info(" - input: " + inputPath);
    sLogger.info(" - output: " + outputPath);
    sLogger.info(" - n: " + n);

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

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

    conf.setInt("mapred.min.split.size", 1024 * 1024 * 1024);
    conf.setInt("n", n);

    FileInputFormat.addInputPath(conf, new Path(inputPath));
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(TextOutputFormat.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(FloatWritable.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(FloatWritable.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);

    return 0;
}

From source file:edu.umd.cloud9.pagerank.PartitionGraph.java

License:Apache License

public int run(String[] args) throws IOException {
    if (args.length != 5) {
        printUsage();//from   ww w . j  a v  a2  s. c  om
        return -1;
    }

    String inPath = args[0];
    String outPath = args[1];
    int numParts = Integer.parseInt(args[2]);
    boolean useRange = Integer.parseInt(args[3]) != 0;
    int nodeCount = Integer.parseInt(args[4]);

    sLogger.info("Tool name: PartitionGraph");
    sLogger.info(" - inputDir: " + inPath);
    sLogger.info(" - outputDir: " + outPath);
    sLogger.info(" - numPartitions: " + numParts);
    sLogger.info(" - useRange?: " + useRange);
    sLogger.info(" - nodeCnt: " + nodeCount);

    JobConf conf = new JobConf(PartitionGraph.class);

    conf.setJobName("Partition Graph " + numParts);
    conf.setNumReduceTasks(numParts);

    conf.setInt("mapred.min.split.size", 1024 * 1024 * 1024);
    conf.set("mapred.child.java.opts", "-Xmx2048m");
    conf.setInt("NodeCount", nodeCount);

    FileInputFormat.setInputPaths(conf, new Path(inPath));
    FileOutputFormat.setOutputPath(conf, new Path(outPath));

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(PageRankNode.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(PageRankNode.class);

    conf.setMapperClass(MapClass.class);
    conf.setReducerClass(ReduceClass.class);

    conf.setSpeculativeExecution(false);

    if (useRange) {
        conf.setPartitionerClass(RangePartitioner.class);
    }

    FileSystem.get(conf).delete(new Path(outPath), true);

    JobClient.runJob(conf);

    return 0;
}

From source file:edu.umd.cloud9.pagerank.RunPageRankBasic.java

License:Apache License

private float phase1(String path, int i, int j, int n, boolean useCombiner, boolean useInmapCombiner,
        boolean useRange) throws IOException {
    JobConf conf = new JobConf(RunPageRankBasic.class);

    String in = path + "/iter" + sFormat.format(i);
    String out = path + "/iter" + sFormat.format(j) + "t";
    String outm = out + "-mass";

    // we need to actually count the number of part files to get the number
    // of partitions (because the directory might contain _log)
    int numPartitions = 0;
    for (FileStatus s : FileSystem.get(conf).listStatus(new Path(in))) {
        if (s.getPath().getName().contains("part-"))
            numPartitions++;//from  w w  w  . j a  v  a  2 s  .  c o  m
    }

    sLogger.info("PageRank: iteration " + j + ": Phase1");
    sLogger.info(" - input: " + in);
    sLogger.info(" - output: " + out);
    sLogger.info(" - nodeCnt: " + n);
    sLogger.info(" - useCombiner: " + useCombiner);
    sLogger.info(" - useInmapCombiner: " + useInmapCombiner);
    sLogger.info(" - useRange: " + useRange);
    sLogger.info("computed number of partitions: " + numPartitions);

    int numMapTasks = numPartitions;
    int numReduceTasks = numPartitions;

    conf.setJobName("PageRank:Basic:iteration" + j + ":Phase1");
    conf.setInt("NodeCount", n);

    conf.setNumMapTasks(numMapTasks);
    conf.setNumReduceTasks(numReduceTasks);

    conf.setInt("mapred.min.split.size", 1024 * 1024 * 1024);
    conf.set("mapred.child.java.opts", "-Xmx2048m");

    conf.set("PageRankMassPath", outm);

    FileInputFormat.setInputPaths(conf, new Path(in));
    FileOutputFormat.setOutputPath(conf, new Path(out));

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(PageRankNode.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(PageRankNode.class);

    if (useInmapCombiner) {
        conf.setMapperClass(MapWithInMapperCombiningClass.class);
    } else {
        conf.setMapperClass(MapClass.class);
    }

    if (useCombiner) {
        conf.setCombinerClass(CombineClass.class);
    }

    if (useRange) {
        conf.setPartitionerClass(RangePartitioner.class);
    }

    conf.setReducerClass(ReduceClass.class);

    conf.setSpeculativeExecution(false);

    FileSystem.get(conf).delete(new Path(out), true);
    FileSystem.get(conf).delete(new Path(outm), true);

    JobClient.runJob(conf);

    float mass = Float.NEGATIVE_INFINITY;
    FileSystem fs = FileSystem.get(conf);
    for (FileStatus f : fs.listStatus(new Path(outm))) {
        FSDataInputStream fin = fs.open(f.getPath());
        mass = sumLogProbs(mass, fin.readFloat());
        fin.close();
    }

    return mass;
}

From source file:edu.umd.cloud9.pagerank.RunPageRankBasic.java

License:Apache License

private void phase2(String path, int i, int j, int n, float missing) throws IOException {
    JobConf conf = new JobConf(RunPageRankBasic.class);

    sLogger.info("missing PageRank mass: " + missing);
    sLogger.info("number of nodes: " + n);

    String in = path + "/iter" + sFormat.format(j) + "t";
    String out = path + "/iter" + sFormat.format(j);

    sLogger.info("PageRank: iteration " + j + ": Phase2");
    sLogger.info(" - input: " + in);
    sLogger.info(" - output: " + out);

    int numMapTasks = FileSystem.get(conf).listStatus(new Path(in)).length;
    int numReduceTasks = 0;

    conf.setJobName("PageRank:Basic:iteration" + j + ":Phase2");
    conf.setInt("mapred.min.split.size", 1024 * 1024 * 1024);
    conf.setFloat("MissingMass", (float) missing);
    conf.setInt("NodeCount", n);

    conf.setNumMapTasks(numMapTasks);/* w w w. jav  a  2  s  .c o m*/
    conf.setNumReduceTasks(numReduceTasks);

    FileInputFormat.setInputPaths(conf, new Path(in));
    FileOutputFormat.setOutputPath(conf, new Path(out));

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(PageRankNode.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(PageRankNode.class);

    conf.setMapperClass(MapPageRankMassDistributionClass.class);
    conf.setCombinerClass(IdentityReducer.class);
    conf.setReducerClass(IdentityReducer.class);

    FileSystem.get(conf).delete(new Path(out), true);

    JobClient.runJob(conf);
}

From source file:edu.umd.cloud9.pagerank.RunPageRankSchimmy.java

License:Apache License

private float phase1(String path, int i, int j, int n, boolean useCombiner, boolean useInmapCombiner,
        boolean useRange) throws IOException {
    JobConf conf = new JobConf(RunPageRankBasic.class);

    String in = path + "/iter" + sFormat.format(i);
    String out = path + "/iter" + sFormat.format(j) + "t";
    String outm = out + "-mass";

    FileSystem fs = FileSystem.get(conf);

    // we need to actually count the number of part files to get the number
    // of partitions (because the directory might contain _log)
    int numPartitions = 0;
    for (FileStatus s : FileSystem.get(conf).listStatus(new Path(in))) {
        if (s.getPath().getName().contains("part-"))
            numPartitions++;/*from   ww w  .j a v  a 2 s .  c o m*/
    }

    conf.setInt("NodeCount", n);

    Partitioner p = null;

    if (useRange) {
        p = new RangePartitioner<IntWritable, Writable>();
        p.configure(conf);
    } else {
        p = new HashPartitioner<WritableComparable, Writable>();
    }

    // this is really annoying: the mapping between the partition numbers on
    // disk (i.e., part-XXXX) and what partition the file contains (i.e.,
    // key.hash % #reducer) is arbitrary... so this means that we need to
    // open up each partition, peek inside to find out.
    IntWritable key = new IntWritable();
    PageRankNode value = new PageRankNode();
    FileStatus[] status = fs.listStatus(new Path(in));

    StringBuilder sb = new StringBuilder();

    for (FileStatus f : status) {
        if (f.getPath().getName().contains("_logs"))
            continue;

        SequenceFile.Reader reader = new SequenceFile.Reader(fs, f.getPath(), conf);

        reader.next(key, value);
        int np = p.getPartition(key, value, numPartitions);
        reader.close();

        sLogger.info(f.getPath() + "\t" + np);
        sb.append(np + "=" + f.getPath() + "\t");
    }

    sLogger.info(sb.toString().trim());

    sLogger.info("PageRankSchimmy: iteration " + j + ": Phase1");
    sLogger.info(" - input: " + in);
    sLogger.info(" - output: " + out);
    sLogger.info(" - nodeCnt: " + n);
    sLogger.info(" - useCombiner: " + useCombiner);
    sLogger.info(" - useInmapCombiner: " + useInmapCombiner);
    sLogger.info(" - numPartitions: " + numPartitions);
    sLogger.info(" - useRange: " + useRange);
    sLogger.info("computed number of partitions: " + numPartitions);

    int numMapTasks = numPartitions;
    int numReduceTasks = numPartitions;

    conf.setJobName("PageRankSchimmy:iteration" + j + ":Phase1");

    conf.setNumMapTasks(numMapTasks);
    conf.setNumReduceTasks(numReduceTasks);

    conf.setInt("mapred.min.split.size", 1024 * 1024 * 1024);
    conf.set("mapred.child.java.opts", "-Xmx2048m");

    conf.set("PageRankMassPath", outm);
    conf.set("BasePath", in);
    conf.set("PartitionMapping", sb.toString().trim());

    FileInputFormat.setInputPaths(conf, new Path(in));
    FileOutputFormat.setOutputPath(conf, new Path(out));

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(FloatWritable.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(PageRankNode.class);

    if (useInmapCombiner) {
        conf.setMapperClass(MapWithInMapperCombiningClass.class);
    } else {
        conf.setMapperClass(MapClass.class);
    }

    if (useCombiner) {
        conf.setCombinerClass(CombineClass.class);
    }

    if (useRange) {
        conf.setPartitionerClass(RangePartitioner.class);
    }

    conf.setReducerClass(ReduceClass.class);

    conf.setSpeculativeExecution(false);

    FileSystem.get(conf).delete(new Path(out), true);
    FileSystem.get(conf).delete(new Path(outm), true);

    JobClient.runJob(conf);

    float mass = Float.NEGATIVE_INFINITY;
    for (FileStatus f : fs.listStatus(new Path(outm))) {
        FSDataInputStream fin = fs.open(f.getPath());
        mass = sumLogProbs(mass, fin.readFloat());
        fin.close();
    }

    return mass;
}

From source file:edu.umd.cloud9.pagerank.RunPageRankSchimmy.java

License:Apache License

private void phase2(String path, int i, int j, int n, float missing) throws IOException {
    JobConf conf = new JobConf(RunPageRankBasic.class);

    sLogger.info("missing PageRank mass: " + missing);
    sLogger.info("number of nodes: " + n);

    String in = path + "/iter" + sFormat.format(j) + "t";
    String out = path + "/iter" + sFormat.format(j);

    sLogger.info("PageRankSchimmy: iteration " + j + ": Phase2");
    sLogger.info(" - input: " + in);
    sLogger.info(" - output: " + out);

    int numMapTasks = FileSystem.get(conf).listStatus(new Path(in)).length;
    int numReduceTasks = 0;

    conf.setJobName("PageRankSchimmy:iteration" + j + ":Phase2");
    conf.setInt("mapred.min.split.size", 1024 * 1024 * 1024);

    conf.setNumMapTasks(numMapTasks);//from  w  w  w  .java  2  s . c  o m
    conf.setNumReduceTasks(numReduceTasks);

    FileInputFormat.setInputPaths(conf, new Path(in));
    FileOutputFormat.setOutputPath(conf, new Path(out));

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(PageRankNode.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(PageRankNode.class);

    conf.setMapperClass(MapPageRankMassDistributionClass.class);
    conf.setCombinerClass(IdentityReducer.class);
    conf.setReducerClass(IdentityReducer.class);

    conf.setFloat("MissingMass", (float) missing);
    conf.setInt("NodeCount", n);

    FileSystem.get(conf).delete(new Path(out), true);

    JobClient.runJob(conf);
}

From source file:edu.umd.cloud9.webgraph.BuildReverseWebGraph.java

License:Apache License

public int runTool() throws Exception {
    JobConf conf = new JobConf(getConf(), BuildReverseWebGraph.class);
    FileSystem fs = FileSystem.get(conf);

    int numMappers = conf.getInt("Cloud9.Mappers", 1);
    int numReducers = conf.getInt("Cloud9.Reducers", 200);

    String inputPath = conf.get("Cloud9.InputPath");
    String outputPath = conf.get("Cloud9.OutputPath");

    conf.setJobName("ReverseWebGraph");
    conf.set("mapred.child.java.opts", "-Xmx2048m");
    conf.setInt("mapred.task.timeout", 60000000);
    conf.set("mapreduce.map.memory.mb", "2048");
    conf.set("mapreduce.map.java.opts", "-Xmx2048m");
    conf.set("mapreduce.reduce.memory.mb", "2048");
    conf.set("mapreduce.reduce.java.opts", "-Xmx2048m");
    conf.set("mapreduce.task.timeout", "60000000");

    conf.setNumMapTasks(numMappers);/*  w  ww. j a  va2s .c om*/
    conf.setNumReduceTasks(numReducers);
    conf.setMapperClass(IdentityMapper.class);
    conf.setReducerClass(Reduce.class);
    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(ArrayListWritable.class);
    conf.setMapOutputKeyClass(Text.class);
    conf.setMapOutputValueClass(ArrayListWritable.class);

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    SequenceFileOutputFormat.setCompressOutput(conf, true);
    SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);

    SequenceFileInputFormat.setInputPaths(conf, inputPath);
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));

    LOG.info("BuildReverseWebGraph");
    LOG.info(" - input path: " + inputPath);
    LOG.info(" - output path: " + outputPath);

    if (!fs.exists(new Path(outputPath))) {
        JobClient.runJob(conf);
    } else {
        LOG.info(outputPath + " already exists! Skipping this step...");
    }

    return 0;
}

From source file:edu.umd.cloud9.webgraph.BuildWebGraph.java

License:Apache License

public int runTool() throws Exception {
    JobConf conf = new JobConf(getConf(), BuildWebGraph.class);
    FileSystem fs = FileSystem.get(conf);

    int numMappers = conf.getInt("Cloud9.Mappers", 1);
    int numReducers = conf.getInt("Cloud9.Reducers", 200);

    String inputPath = conf.get("Cloud9.InputPath");
    String outputPath = conf.get("Cloud9.OutputPath");

    conf.setJobName("ConstructWebGraph");
    conf.set("mapred.child.java.opts", "-Xmx2048m");
    conf.setInt("mapred.task.timeout", 60000000);
    conf.set("mapreduce.map.memory.mb", "2048");
    conf.set("mapreduce.map.java.opts", "-Xmx2048m");
    conf.set("mapreduce.reduce.memory.mb", "2048");
    conf.set("mapreduce.reduce.java.opts", "-Xmx2048m");
    conf.set("mapreduce.task.timeout", "60000000");

    conf.setNumMapTasks(numMappers);/*from   www . j av a 2 s .  c o m*/
    conf.setNumReduceTasks(numReducers);

    conf.setMapperClass(Map.class);
    conf.setReducerClass(Reduce.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(ArrayListWritable.class);

    conf.setMapOutputKeyClass(IntWritable.class);
    conf.setMapOutputValueClass(ArrayListWritable.class);

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    SequenceFileOutputFormat.setCompressOutput(conf, true);
    SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);

    SequenceFileInputFormat.setInputPaths(conf, inputPath);
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));

    LOG.info("BuildWebGraph");
    LOG.info(" - input path: " + inputPath);
    LOG.info(" - output path: " + outputPath);

    if (!fs.exists(new Path(outputPath))) {
        JobClient.runJob(conf);
    } else {
        LOG.info(outputPath + " already exists! Skipping this step...");
    }

    return 0;
}

From source file:edu.umd.cloud9.webgraph.ClueExtractLinks.java

License:Apache License

public int runTool() throws Exception {
    JobConf conf = new JobConf(getConf(), ClueExtractLinks.class);
    FileSystem fs = FileSystem.get(conf);

    int numMappers = conf.getInt("Cloud9.Mappers", 1);
    int numReducers = conf.getInt("Cloud9.Reducers", 200);

    String inputPath = conf.get("Cloud9.InputPath");
    String outputPath = conf.get("Cloud9.OutputPath");
    String mappingFile = conf.get("Cloud9.DocnoMappingFile");

    if (!fs.exists(new Path(mappingFile))) {
        throw new RuntimeException("Error: Docno mapping data file " + mappingFile + " doesn't exist!");
    }/*from w  w w.j  a  v  a2  s. c  o m*/

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

    conf.setJobName("ClueExtractLinks");
    conf.set("mapred.child.java.opts", "-Xmx2048m");
    conf.setInt("mapred.task.timeout", 60000000);
    conf.set("mapreduce.map.memory.mb", "2048");
    conf.set("mapreduce.map.java.opts", "-Xmx2048m");
    conf.set("mapreduce.reduce.memory.mb", "2048");
    conf.set("mapreduce.reduce.java.opts", "-Xmx2048m");
    conf.set("mapreduce.task.timeout", "60000000");

    conf.setNumMapTasks(numMappers);
    conf.setNumReduceTasks(numReducers);
    // TODO: to read!!
    conf.setMapperClass(Map.class);
    conf.setCombinerClass(Reduce.class);
    conf.setReducerClass(Reduce.class);
    conf.setOutputKeyClass(Text.class);
    conf.setOutputValueClass(ArrayListWritable.class);
    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);
    SequenceFileOutputFormat.setCompressOutput(conf, true);
    SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);
    SequenceFileInputFormat.setInputPaths(conf, inputPath);
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));

    LOG.info("ClueExtractLinks");
    LOG.info(" - input path: " + inputPath);
    LOG.info(" - output path: " + outputPath);
    LOG.info(" - mapping file: " + mappingFile);
    LOG.info(" - include internal links? " + conf.getBoolean("Cloud9.IncludeInternalLinks", false));

    if (!fs.exists(new Path(outputPath))) {
        JobClient.runJob(conf);
    } else {
        LOG.info(outputPath + " already exists! Skipping this step...");
    }
    return 0;
}

From source file:edu.umd.cloud9.webgraph.CollectHostnames.java

License:Apache License

public int runTool() throws Exception {

    JobConf conf = new JobConf(getConf(), CollectHostnames.class);
    FileSystem fs = FileSystem.get(conf);

    int numMappers = conf.getInt("Cloud9.Mappers", 1);
    int numReducers = conf.getInt("Cloud9.Reducers", 200);

    String inputPath = conf.get("Cloud9.InputPath");
    String outputPath = conf.get("Cloud9.OutputPath");

    conf.setJobName("CollectHostnames");
    conf.set("mapred.child.java.opts", "-Xmx4096m");
    conf.setInt("mapred.task.timeout", 60000000);

    conf.setNumMapTasks(numMappers);//from  w  ww .  j av a2 s  .  c om
    conf.setNumReduceTasks(numReducers);

    conf.setMapperClass(Map.class);
    conf.setPartitionerClass(Partition.class);
    conf.setReducerClass(Reduce.class);

    conf.setOutputKeyClass(IntWritable.class);
    conf.setOutputValueClass(ArrayListWritable.class);

    conf.setMapOutputKeyClass(PairOfIntString.class);
    conf.setMapOutputValueClass(IntWritable.class);

    conf.setInputFormat(SequenceFileInputFormat.class);
    conf.setOutputFormat(SequenceFileOutputFormat.class);

    SequenceFileOutputFormat.setCompressOutput(conf, true);
    SequenceFileOutputFormat.setOutputCompressionType(conf, SequenceFile.CompressionType.BLOCK);

    SequenceFileInputFormat.setInputPaths(conf, inputPath);
    FileOutputFormat.setOutputPath(conf, new Path(outputPath));

    sLogger.info("PropagateHostname");
    sLogger.info(" - input path: " + inputPath);
    sLogger.info(" - output path: " + outputPath);

    if (!fs.exists(new Path(outputPath))) {
        JobClient.runJob(conf);
    } else {
        sLogger.info(outputPath + " already exists! Skipping this step...");
    }

    return 0;
}