Example usage for org.apache.hadoop.mapreduce Job setMapperClass

List of usage examples for org.apache.hadoop.mapreduce Job setMapperClass

  1. HOME
  2. Java
  3. org.apache.hadoop
  4. org.apache.hadoop.mapreduce.*
  5. Job
  6. setMapperClass

Introduction

In this page you can find the example usage for org.apache.hadoop.mapreduce Job setMapperClass.

Prototype

public void setMapperClass(Class<? extends Mapper> cls) throws IllegalStateException

Source Link

Document

Set the Mapper for the job.

Usage

From source file:com.avira.couchdoop.jobs.CouchbaseViewToHBaseImporter.java

License:Apache License

public Job configureJob(Configuration conf, String outputTable) throws IOException {
    conf.setInt("mapreduce.map.failures.maxpercent", 5);
    conf.setInt("mapred.max.map.failures.percent", 5);
    conf.setInt("mapred.max.tracker.failures", 20);

    Job job = Job.getInstance(conf);
    job.setJarByClass(CouchbaseViewToHBaseImporter.class);

    // Input//from  w w w  . j a v  a  2  s . c  om
    job.setInputFormatClass(CouchbaseViewInputFormat.class);

    // Mapper
    job.setMapperClass(CouchbaseViewToHBaseMapper.class);

    // Reducer
    job.setNumReduceTasks(0);

    // Output
    TableMapReduceUtil.initTableReducerJob(outputTable, IdentityTableReducer.class, job);

    return job;
}

From source file:com.awcoleman.BouncyCastleGenericCDRHadoop.BasicDriverMapReduce.java

License:Apache License

public int run(String[] args) throws Exception {

    if (args.length < 2) {
        System.out.println("Missing input and output filenames. Exiting.");
        System.exit(1);//from  w  w  w .  j a va2s.  com
    }

    Job job = new Job(super.getConf());
    job.setJarByClass(BasicDriverMapReduce.class);
    job.setJobName("BasicDriver1");
    job.setMapperClass(BasicMapper.class);
    job.setReducerClass(BasicReducer.class);
    job.setOutputKeyClass(Text.class);
    job.setOutputValueClass(LongWritable.class);
    job.setInputFormatClass(RawFileAsBinaryInputFormat.class);
    FileInputFormat.addInputPath(job, new Path(args[0]));
    FileOutputFormat.setOutputPath(job, new Path(args[1]));

    return job.waitForCompletion(true) ? 0 : 1;
}

From source file:com.awcoleman.BouncyCastleGenericCDRHadoopWithWritable.BasicDriverMapReduce.java

License:Apache License

public int run(String[] args) throws Exception {

    if (args.length < 2) {
        System.out.println("Missing input and output filenames. Exiting.");
        System.exit(1);/*from  w  w w.j  a  v  a2s.co m*/
    }

    @SuppressWarnings("deprecation")
    Job job = new Job(super.getConf());
    job.setJarByClass(BasicDriverMapReduce.class);
    job.setJobName("BasicDriverMapReduce");
    job.setMapperClass(BasicMapper.class);
    job.setReducerClass(BasicReducer.class);
    job.setMapOutputKeyClass(Text.class);
    job.setMapOutputValueClass(CallDetailRecord.class);
    job.setOutputKeyClass(Text.class);
    job.setOutputValueClass(LongWritable.class);
    job.setInputFormatClass(RawFileAsBinaryInputFormat.class);
    FileInputFormat.addInputPath(job, new Path(args[0]));
    FileOutputFormat.setOutputPath(job, new Path(args[1]));

    return job.waitForCompletion(true) ? 0 : 1;
}

From source file:com.awcoleman.ExampleJobSummaryLogWithOutput.BinRecToAvroRecDriver.java

License:Apache License

public int run(String[] args) throws Exception {

    String input = null;/*from w w  w. j  a  v  a2s  .com*/
    String output = null;

    if (args.length < 2) {
        System.err.printf("Usage: %s <input> <output>\n", this.getClass().getSimpleName());
        return -1;
    } else {
        input = args[0];
        output = args[1];
    }

    Job job = Job.getInstance(getConf(), "BinRecToAvroRecDriver");
    Configuration conf = job.getConfiguration();

    //Add job log to hold Driver logging (and any summary info about the dataset,job, or counters we want to write)
    String fapath = createTempFileAppender(job);

    //get schema
    Schema outSchema = ReflectData.get().getSchema(com.awcoleman.examples.avro.BinRecForPartitions.class);
    job.getConfiguration().set("outSchema", outSchema.toString());

    //Job conf settings
    job.setJarByClass(BinRecToAvroRecDriver.class);
    job.setMapperClass(Map.class);
    job.setReducerClass(Reduce.class);
    job.setInputFormatClass(BinRecInputFormat.class);
    job.setOutputFormatClass(AvroKeyOutputFormat.class);
    AvroJob.setOutputKeySchema(job, outSchema);

    AvroJob.setMapOutputKeySchema(job, Schema.create(Schema.Type.STRING));
    AvroJob.setMapOutputValueSchema(job, outSchema);

    //Job output compression
    FileOutputFormat.setCompressOutput(job, true);
    job.getConfiguration().set(AvroJob.CONF_OUTPUT_CODEC, DataFileConstants.DEFLATE_CODEC);

    //Input and Output Paths
    FileInputFormat.setInputPaths(job, new Path(input));
    Path outPath = new Path(output);
    FileOutputFormat.setOutputPath(job, outPath);
    outPath.getFileSystem(conf).delete(outPath, true);

    boolean jobCompletionStatus = job.waitForCompletion(true);

    //Print Custom Counters before exiting
    Counters counters = job.getCounters();
    for (MYJOB_CNTRS customCounter : MYJOB_CNTRS.values()) {
        Counter thisCounter = counters.findCounter(customCounter);
        System.out.println("Custom Counter " + customCounter + "=" + thisCounter.getValue());
    }

    long mycnt1 = job.getCounters()
            .findCounter("com.awcoleman.TestingGettingContainerLogger.BinRecToAvroRecDriver$MYJOB_CNTRS",
                    "MYCNT1")
            .getValue();
    long mycnt2 = job.getCounters()
            .findCounter("com.awcoleman.TestingGettingContainerLogger.BinRecToAvroRecDriver$MYJOB_CNTRS",
                    "MYCNT2")
            .getValue();
    long mycnt3 = job.getCounters()
            .findCounter("com.awcoleman.TestingGettingContainerLogger.BinRecToAvroRecDriver$MYJOB_CNTRS",
                    "MYCNT3")
            .getValue();

    long myfakekpi = mycnt1 - mycnt2;

    String msgMyfakekpi = "The Fake KPI of the Dataset: " + String.format("%,d", myfakekpi);
    System.out.println(msgMyfakekpi);
    logger.info(msgMyfakekpi);

    //Finished, so move job log to HDFS in _log dir, clean
    copyTempFileAppenderToHDFSOutpath(job, fapath, output);

    return jobCompletionStatus ? 0 : 1;
}

From source file:com.bah.applefox.main.plugins.fulltextindex.FTLoader.java

License:Apache License

/**
 * run takes the comandline args as arguments (in this case from a
 * configuration file), creates a new job, configures it, initiates it,
 * waits for completion, and returns 0 if it is successful (1 if it is not)
 * /*from  w ww  . ja v a  2  s .c  o m*/
 * @param args
 *            the commandline arguments (in this case from a configuration
 *            file)
 * 
 * @return 0 if the job ran successfully and 1 if it isn't
 */
public int run(String[] args) throws Exception {
    try {
        // Initialize variables
        FTLoader.articleFile = args[8];
        FTLoader.maxNGrams = Integer.parseInt(args[9]);
        FTLoader.stopWords = getStopWords();
        FTLoader.dTable = args[10];
        FTLoader.urlCheckedTable = args[11];
        FTLoader.divsFile = args[20];
        FTLoader.exDivs = getExDivs();

        // Give the job a name
        String jobName = this.getClass().getSimpleName() + "_" + System.currentTimeMillis();

        // Create job and set the jar
        Job job = new Job(getConf(), jobName);
        job.setJarByClass(this.getClass());

        String urlTable = args[5];

        job.setInputFormatClass(AccumuloInputFormat.class);
        InputFormatBase.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
        InputFormatBase.setInputInfo(job.getConfiguration(), args[2], args[3].getBytes(), urlTable,
                new Authorizations());

        job.setMapperClass(MapperClass.class);
        job.setMapOutputKeyClass(Key.class);
        job.setMapOutputValueClass(Value.class);

        job.setReducerClass(ReducerClass.class);
        job.setNumReduceTasks(Integer.parseInt(args[4]));

        job.setOutputFormatClass(AccumuloOutputFormat.class);
        job.setOutputKeyClass(Key.class);
        job.setOutputValueClass(Value.class);

        AccumuloOutputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
        AccumuloOutputFormat.setOutputInfo(job.getConfiguration(), args[2], args[3].getBytes(), true, urlTable);

        job.waitForCompletion(true);

        return job.isSuccessful() ? 0 : 1;
    } catch (IOException e) {
        if (e.getMessage() != null) {
            log.error(e.getMessage());
        } else {
            log.error(e.getStackTrace());
        }
    } catch (InterruptedException e) {
        if (e.getMessage() != null) {
            log.error(e.getMessage());
        } else {
            log.error(e.getStackTrace());
        }
    } catch (ClassNotFoundException e) {
        if (e.getMessage() != null) {
            log.error(e.getMessage());
        } else {
            log.error(e.getStackTrace());
        }
    }
    return 1;
}

From source file:com.bah.applefox.main.plugins.imageindex.ImageLoader.java

License:Apache License

/**
 * run takes the comandline args as arguments (in this case from a
 * configuration file), creates a new job, configures it, initiates it,
 * waits for completion, and returns 0 if it is successful (1 if it is not)
 * /*ww  w  .ja  v  a2s.  c  o m*/
 * @param args
 *            the commandline arguments (in this case from a configuration
 *            file)
 * 
 * @return 0 if the job ran successfully and 1 if it isn't
 */
public int run(String[] args) throws Exception {

    checkedImages = args[18];
    hashTable = args[17];
    tagTable = args[19];
    divsFile = args[20];
    UserAgent = args[6];

    // Create the table
    AccumuloUtils.setSplitSize(args[23]);
    AccumuloUtils.connectBatchWrite(checkedImages).close();

    // Give the job a name
    String jobName = this.getClass().getSimpleName() + "_" + System.currentTimeMillis();

    // Create the job and set its jar
    Job job = new Job(getConf(), jobName);
    job.setJarByClass(this.getClass());

    // Set the url table to read from
    String urlTable = args[5];

    job.setInputFormatClass(AccumuloInputFormat.class);
    InputFormatBase.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    InputFormatBase.setInputInfo(job.getConfiguration(), args[2], args[3].getBytes(), urlTable,
            new Authorizations());

    job.setMapperClass(MapperClass.class);
    job.setMapOutputKeyClass(Key.class);
    job.setMapOutputValueClass(Value.class);

    job.setNumReduceTasks(Integer.parseInt(args[4]));

    job.setReducerClass(ReducerClass.class);

    job.setOutputFormatClass(AccumuloOutputFormat.class);
    job.setOutputKeyClass(Key.class);
    job.setOutputValueClass(Value.class);
    AccumuloOutputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    AccumuloOutputFormat.setOutputInfo(job.getConfiguration(), args[2], args[3].getBytes(), true, urlTable);

    AccumuloUtils.setSplitSize(args[22]);

    job.waitForCompletion(true);

    return job.isSuccessful() ? 0 : 1;
}

From source file:com.bah.applefox.main.plugins.pageranking.utilities.CountURLs.java

License:Apache License

public int run(String[] args) throws Exception {

    String jobName = this.getClass().getSimpleName() + "_" + System.currentTimeMillis();

    Job job = new Job(getConf(), jobName);
    job.setJarByClass(this.getClass());

    mappedInput = args[12] + "From";

    job.setInputFormatClass(AccumuloInputFormat.class);
    InputFormatBase.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    InputFormatBase.setInputInfo(job.getConfiguration(), args[2], args[3].getBytes(), mappedInput,
            new Authorizations());

    job.setMapperClass(MapperClass.class);
    job.setMapOutputKeyClass(Key.class);
    job.setMapOutputValueClass(Value.class);

    job.setReducerClass(ReducerClass.class);
    job.setOutputFormatClass(AccumuloOutputFormat.class);
    job.setOutputKeyClass(Key.class);
    job.setOutputValueClass(Value.class);
    AccumuloOutputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    AccumuloOutputFormat.setOutputInfo(job.getConfiguration(), args[2], args[3].getBytes(), true, args[15]);

    job.waitForCompletion(true);/*from ww  w .j a v a  2 s .c  om*/

    return job.isSuccessful() ? 0 : 1;
}

From source file:com.bah.applefox.main.plugins.pageranking.utilities.DampenTable.java

License:Apache License

public int run(String[] args) throws Exception {

    String jobName = this.getClass().getSimpleName() + "_" + System.currentTimeMillis();

    Job job = new Job(getConf(), jobName);
    job.setJarByClass(this.getClass());

    tablePrefix = args[13];//from   w  ww  .java 2s  .  c  o m
    dampeningFactor = Double.parseDouble(args[14]);

    job.setInputFormatClass(AccumuloInputFormat.class);
    AccumuloInputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    AccumuloInputFormat.setInputInfo(job.getConfiguration(), args[2], args[3].getBytes(), tablePrefix + "New",
            new Authorizations());

    job.setMapperClass(MapperClass.class);
    job.setMapOutputKeyClass(Key.class);
    job.setMapOutputValueClass(Value.class);

    job.setReducerClass(ReducerClass.class);
    job.setOutputFormatClass(AccumuloOutputFormat.class);
    job.setOutputKeyClass(Key.class);
    job.setOutputValueClass(Value.class);
    AccumuloOutputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    AccumuloOutputFormat.setOutputInfo(job.getConfiguration(), args[2], args[3].getBytes(), true,
            tablePrefix + "New");

    job.waitForCompletion(true);

    return job.isSuccessful() ? 0 : 1;
}

From source file:com.bah.applefox.main.plugins.pageranking.utilities.InitializePRTables.java

License:Apache License

public int run(String[] args) throws Exception {

    tablePrefix = args[13];/*from   w  ww .ja  v  a2 s. c  o m*/

    String jobName = this.getClass().getSimpleName() + "_" + System.currentTimeMillis();

    Job job = new Job(getConf(), jobName);
    job.setJarByClass(this.getClass());

    job.setInputFormatClass(AccumuloInputFormat.class);
    InputFormatBase.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    InputFormatBase.setInputInfo(job.getConfiguration(), args[2], args[3].getBytes(), args[12] + "To",
            new Authorizations());

    job.setMapperClass(MapperClass.class);
    job.setMapOutputKeyClass(Key.class);
    job.setMapOutputValueClass(Value.class);

    job.setReducerClass(ReducerClass.class);
    job.setOutputFormatClass(AccumuloOutputFormat.class);
    job.setOutputKeyClass(Key.class);
    job.setOutputValueClass(Value.class);
    AccumuloOutputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    AccumuloOutputFormat.setOutputInfo(job.getConfiguration(), args[2], args[3].getBytes(), true,
            tablePrefix + "Old");

    AccumuloUtils.connectBatchWrite(tablePrefix + "New");

    job.waitForCompletion(true);

    return job.isSuccessful() ? 0 : 1;
}

From source file:com.bah.applefox.main.plugins.pageranking.utilities.MRPageRanking.java

License:Apache License

public int run(String[] args) throws Exception {

    String jobName = this.getClass().getSimpleName() + "_" + System.currentTimeMillis();

    Job job = new Job(getConf(), jobName);
    job.setJarByClass(this.getClass());

    tablePrefix = args[13];// w w w .  ja v a2s.co m
    outboundLinks = args[15];

    job.setInputFormatClass(AccumuloInputFormat.class);
    AccumuloInputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);

    AccumuloInputFormat.setInputInfo(job.getConfiguration(), args[2], args[3].getBytes(), args[12] + "To",
            new Authorizations());

    job.setMapperClass(MapperClass.class);
    job.setMapOutputKeyClass(Key.class);
    job.setMapOutputValueClass(Value.class);

    job.setReducerClass(ReducerClass.class);
    job.setOutputFormatClass(AccumuloOutputFormat.class);
    job.setOutputKeyClass(Key.class);
    job.setOutputValueClass(Value.class);
    AccumuloOutputFormat.setZooKeeperInstance(job.getConfiguration(), args[0], args[1]);
    AccumuloOutputFormat.setOutputInfo(job.getConfiguration(), args[2], args[3].getBytes(), true,
            tablePrefix + "New");

    job.waitForCompletion(true);

    return job.isSuccessful() ? 0 : 1;
}