List of usage examples for org.apache.hadoop.mapred JobConf set
public void set(String name, String value)
value of the name property. From source file:Corrector.Config.java
License:Apache License
public static void initializeConfiguration(JobConf conf) { validateConfiguration();/*from ww w . jav a 2s.c o m*/ conf.setNumMapTasks(HADOOP_MAPPERS); conf.setNumReduceTasks(HADOOP_REDUCERS); conf.set("mapred.child.java.opts", HADOOP_JAVAOPTS); conf.set("mapred.task.timeout", Long.toString(HADOOP_TIMEOUT)); conf.setLong("LOCALNODES", HADOOP_LOCALNODES); conf.setLong("RANDOM_PASS", RANDOM_PASS); conf.setLong("UP_KMER", UP_KMER); conf.setLong("LOW_KMER", LOW_KMER); conf.setLong("K", K); conf.setLong("READLENGTH", READLEN); }
From source file:crunch.MaxTemperature.java
License:Apache License
@Test
@Ignore("Needs more investigation")
public void recordsCoincideWithBlocks() throws IOException {
int recordLength = 1024;
Path input = new Path("input");
createFile(input, 12, recordLength);
JobConf job = new JobConf();
job.set("fs.default.name", fs.getUri().toString());
FileInputFormat.addInputPath(job, input);
InputFormat<LongWritable, Text> inputFormat = job.getInputFormat();
InputSplit[] splits = inputFormat.getSplits(job, job.getNumMapTasks());
assertThat(splits.length, is(3));
checkSplit(splits[0], 0, 4096);//from w w w. j a v a 2 s.co m
checkSplit(splits[1], 4096, 4096);
checkSplit(splits[2], 8192, 4096);
checkRecordReader(inputFormat, splits[0], job, recordLength, 0, 4);
checkRecordReader(inputFormat, splits[1], job, recordLength, 4, 8);
checkRecordReader(inputFormat, splits[2], job, recordLength, 8, 12);
}
From source file:crunch.MaxTemperature.java
License:Apache License
@Test
public void recordsDontCoincideWithBlocks() throws IOException {
int recordLength = 1024 + 512;
Path input = new Path("input");
createFile(input, 8, recordLength);
JobConf job = new JobConf();
job.set("fs.default.name", fs.getUri().toString());
FileInputFormat.addInputPath(job, input);
InputFormat<LongWritable, Text> inputFormat = job.getInputFormat();
InputSplit[] splits = inputFormat.getSplits(job, job.getNumMapTasks());
System.out.println(Arrays.asList(splits));
checkSplit(splits[0], 0, 4096);//w w w . ja va 2 s . co m
checkSplit(splits[1], 4096, 4096);
checkSplit(splits[2], 8192, 4096);
checkRecordReader(inputFormat, splits[0], job, recordLength, 0, 3);
checkRecordReader(inputFormat, splits[1], job, recordLength, 3, 6);
checkRecordReader(inputFormat, splits[2], job, recordLength, 6, 8);
}
From source file:crunch.MaxTemperature.java
License:Apache License
@Test
@Ignore("Needs more investigation")
public void compression() throws IOException {
int recordLength = 1024;
Path input = new Path("input.bz2");
createFile(input, 24, recordLength);
System.out.println(">>>>>>" + fs.getLength(input));
JobConf job = new JobConf();
job.set("fs.default.name", fs.getUri().toString());
FileInputFormat.addInputPath(job, input);
InputFormat<LongWritable, Text> inputFormat = job.getInputFormat();
InputSplit[] splits = inputFormat.getSplits(job, job.getNumMapTasks());
System.out.println(Arrays.asList(splits));
assertThat(splits.length, is(2));
checkSplit(splits[0], 0, 4096);/*from w ww . ja v a2 s . c om*/
checkSplit(splits[1], 4096, 4096);
checkRecordReader(inputFormat, splits[0], job, recordLength, 0, 4);
checkRecordReader(inputFormat, splits[1], job, recordLength, 5, 12);
}
From source file:de.l3s.streamcorpus.mapreduce.TerrierIndexing.java
License:Mozilla Public License
/** Starts the MapReduce indexing. * @param args/* w w w . ja v a 2 s . c om*/ * @throws Exception */ public int run(String[] args) throws Exception { long time = System.currentTimeMillis(); // For the moment: Hard-code the terrier home to quick test System.setProperty("terrier.home", "/home/tuan.tran/executable/StreamCorpusIndexer"); boolean docPartitioned = false; int numberOfReducers = Integer .parseInt(ApplicationSetup.getProperty("terrier.hadoop.indexing.reducers", "26")); final HadoopPlugin.JobFactory jf = HadoopPlugin.getJobFactory("HOD-TerrierIndexing"); if (args.length == 2 && args[0].equals("-p")) { logger.debug("Document-partitioned Mode, " + numberOfReducers + " output indices."); numberOfReducers = Integer.parseInt(args[1]); docPartitioned = true; } else if (args.length == 1 && args[0].equals("--merge")) { if (numberOfReducers > 1) mergeLexiconInvertedFiles(ApplicationSetup.TERRIER_INDEX_PATH, numberOfReducers); else logger.error("No point merging 1 reduce task output"); return 0; } else if (args.length == 0) { logger.debug("Term-partitioned Mode, " + numberOfReducers + " reducers creating one inverted index."); docPartitioned = false; if (numberOfReducers > MAX_REDUCE) { logger.warn("Excessive reduce tasks (" + numberOfReducers + ") in use " + "- SplitEmittedTerm.SETPartitionerLowercaseAlphaTerm can use " + MAX_REDUCE + " at most"); } } /*else { logger.fatal(usage()); return 0; }*/ if (!(CompressionFactory.getCompressionConfiguration("inverted", new String[0], false) instanceof BitCompressionConfiguration)) { logger.error("Sorry, only default BitCompressionConfiguration is supported by HadoopIndexing" + " - you can recompress the inverted index later using IndexRecompressor"); return 0; } if (jf == null) throw new Exception("Could not get JobFactory from HadoopPlugin"); final JobConf conf = jf.newJob(); conf.setJarByClass(TerrierIndexing.class); conf.setJobName("StreamCorpusIndexer: Terrier Indexing"); if (Files.exists(ApplicationSetup.TERRIER_INDEX_PATH) && Index.existsIndex(ApplicationSetup.TERRIER_INDEX_PATH, ApplicationSetup.TERRIER_INDEX_PREFIX)) { logger.fatal("Cannot index while index exists at " + ApplicationSetup.TERRIER_INDEX_PATH + "," + ApplicationSetup.TERRIER_INDEX_PREFIX); return 0; } // boolean blockIndexing = ApplicationSetup.BLOCK_INDEXING; boolean blockIndexing = true; if (blockIndexing) { conf.setMapperClass(Hadoop_BlockSinglePassIndexer.class); conf.setReducerClass(Hadoop_BlockSinglePassIndexer.class); } else { conf.setMapperClass(Hadoop_BasicSinglePassIndexer.class); conf.setReducerClass(Hadoop_BasicSinglePassIndexer.class); } FileOutputFormat.setOutputPath(conf, new Path(ApplicationSetup.TERRIER_INDEX_PATH)); conf.set("indexing.hadoop.prefix", ApplicationSetup.TERRIER_INDEX_PREFIX); conf.setMapOutputKeyClass(SplitEmittedTerm.class); conf.setMapOutputValueClass(MapEmittedPostingList.class); conf.setBoolean("indexing.hadoop.multiple.indices", docPartitioned); if (!conf.get("mapred.job.tracker").equals("local")) { conf.setMapOutputCompressorClass(GzipCodec.class); conf.setCompressMapOutput(true); } else { conf.setCompressMapOutput(false); } conf.setInputFormat(MultiFileCollectionInputFormat.class); conf.setOutputFormat(NullOutputFormat.class); conf.setOutputKeyComparatorClass(SplitEmittedTerm.SETRawComparatorTermSplitFlush.class); conf.setOutputValueGroupingComparator(SplitEmittedTerm.SETRawComparatorTerm.class); conf.setReduceSpeculativeExecution(false); //parse the collection.spec BufferedReader specBR = Files.openFileReader(ApplicationSetup.COLLECTION_SPEC); String line = null; List<Path> paths = new ArrayList<Path>(); while ((line = specBR.readLine()) != null) { if (line.startsWith("#")) continue; paths.add(new Path(line)); } specBR.close(); FileInputFormat.setInputPaths(conf, paths.toArray(new Path[paths.size()])); // not sure if this is effective in YARN conf.setNumMapTasks(2000); // increase the heap usage conf.set("mapreduce.map.memory.mb", "6100"); conf.set("mapred.job.map.memory.mb", "6100"); conf.set("mapreduce.reduce.memory.mb", "6144"); conf.set("mapred.job.reduce.memory.mb", "6144"); conf.set("mapreduce.map.java.opts", "-Xmx6100m"); conf.set("mapred.map.child.java.opts", "-Xmx6100m"); conf.set("mapreduce.reduce.java.opts", "-Xmx6144m"); conf.set("mapred.reduce.child.opts", "-Xmx6144m"); //conf.setBoolean("mapred.used.genericoptionsparser", true) ; // This is the nasty thing in MapReduce v2 and YARN: They always prefer their ancient jars first. Set this on to say you don't like it conf.set("mapreduce.job.user.classpath.first", "true"); // increase the yarn memory to 10 GB conf.set("yarn.nodemanager.resource.memory-mb", "12288"); conf.set("yarn.nodemanager.resource.cpu-vcores", "16"); conf.set("yarn.scheduler.minimum-allocation-mb", "4096"); conf.setNumReduceTasks(numberOfReducers); if (numberOfReducers > 1) { if (docPartitioned) conf.setPartitionerClass(SplitEmittedTerm.SETPartitioner.class); else conf.setPartitionerClass(SplitEmittedTerm.SETPartitionerLowercaseAlphaTerm.class); } else { //for JUnit tests, we seem to need to restore the original partitioner class conf.setPartitionerClass(HashPartitioner.class); } /*JobID jobId = null; boolean ranOK = true; try{ RunningJob rj = JobClient.runJob(conf); jobId = rj.getID(); HadoopUtility.finishTerrierJob(conf); } catch (Exception e) { logger.error("Problem running job", e); e.printStackTrace(); ranOK = false; } if (jobId != null) { deleteTaskFiles(ApplicationSetup.TERRIER_INDEX_PATH, jobId); } */ //if (ranOK) //{ System.out.println("Merging indices"); if (!docPartitioned) { if (numberOfReducers > 1) mergeLexiconInvertedFiles(ApplicationSetup.TERRIER_INDEX_PATH, numberOfReducers); } Hadoop_BasicSinglePassIndexer.finish(ApplicationSetup.TERRIER_INDEX_PATH, docPartitioned ? numberOfReducers : 1, jf); //} System.out.println("Time Taken = " + ((System.currentTimeMillis() - time) / 1000) + " seconds"); jf.close(); return 0; }
From source file:de.l3s.streamcorpus.StreamCorpusIndexing.java
License:Mozilla Public License
/** Starts the MapReduce indexing. * @param args//from www .j av a 2 s . c o m * @throws Exception */ public int run(String[] args) throws Exception { long time = System.currentTimeMillis(); // For the moment: Hard-code the terrier home to quick test System.setProperty("terrier.home", "/home/tuan.tran/executable/StreamCorpusIndexer"); boolean docPartitioned = false; int numberOfReducers = Integer .parseInt(ApplicationSetup.getProperty("terrier.hadoop.indexing.reducers", "26")); final HadoopPlugin.JobFactory jf = HadoopPlugin.getJobFactory("HOD-TerrierIndexing"); if (args.length == 2 && args[0].equals("-p")) { logger.debug("Document-partitioned Mode, " + numberOfReducers + " output indices."); numberOfReducers = Integer.parseInt(args[1]); docPartitioned = true; } else if (args.length == 1 && args[0].equals("--merge")) { if (numberOfReducers > 1) mergeLexiconInvertedFiles(ApplicationSetup.TERRIER_INDEX_PATH, numberOfReducers); else logger.error("No point merging 1 reduce task output"); return 0; } else if (args.length == 0) { logger.debug("Term-partitioned Mode, " + numberOfReducers + " reducers creating one inverted index."); docPartitioned = false; if (numberOfReducers > MAX_REDUCE) { logger.warn("Excessive reduce tasks (" + numberOfReducers + ") in use " + "- SplitEmittedTerm.SETPartitionerLowercaseAlphaTerm can use " + MAX_REDUCE + " at most"); } } /*else { logger.fatal(usage()); return 0; }*/ if (!(CompressionFactory.getCompressionConfiguration("inverted", new String[0], false) instanceof BitCompressionConfiguration)) { logger.error("Sorry, only default BitCompressionConfiguration is supported by HadoopIndexing" + " - you can recompress the inverted index later using IndexRecompressor"); return 0; } if (jf == null) throw new Exception("Could not get JobFactory from HadoopPlugin"); final JobConf conf = jf.newJob(); conf.setJarByClass(StreamCorpusIndexing.class); conf.setJobName("StreamCorpusIndexer: Terrier Indexing"); if (Files.exists(ApplicationSetup.TERRIER_INDEX_PATH) && Index.existsIndex(ApplicationSetup.TERRIER_INDEX_PATH, ApplicationSetup.TERRIER_INDEX_PREFIX)) { logger.fatal("Cannot index while index exists at " + ApplicationSetup.TERRIER_INDEX_PATH + "," + ApplicationSetup.TERRIER_INDEX_PREFIX); return 0; } // boolean blockIndexing = ApplicationSetup.BLOCK_INDEXING; boolean blockIndexing = true; if (blockIndexing) { conf.setMapperClass(Hadoop_BlockSinglePassIndexer.class); conf.setReducerClass(Hadoop_BlockSinglePassIndexer.class); } else { conf.setMapperClass(Hadoop_BasicSinglePassIndexer.class); conf.setReducerClass(Hadoop_BasicSinglePassIndexer.class); } FileOutputFormat.setOutputPath(conf, new Path(ApplicationSetup.TERRIER_INDEX_PATH)); conf.set("indexing.hadoop.prefix", ApplicationSetup.TERRIER_INDEX_PREFIX); conf.setMapOutputKeyClass(SplitEmittedTerm.class); conf.setMapOutputValueClass(MapEmittedPostingList.class); conf.setBoolean("indexing.hadoop.multiple.indices", docPartitioned); if (!conf.get("mapred.job.tracker").equals("local")) { conf.setMapOutputCompressorClass(GzipCodec.class); conf.setCompressMapOutput(true); } else { conf.setCompressMapOutput(false); } conf.setInputFormat(MultiFileCollectionInputFormat.class); conf.setOutputFormat(NullOutputFormat.class); conf.setOutputKeyComparatorClass(SplitEmittedTerm.SETRawComparatorTermSplitFlush.class); conf.setOutputValueGroupingComparator(SplitEmittedTerm.SETRawComparatorTerm.class); conf.setReduceSpeculativeExecution(false); //parse the collection.spec BufferedReader specBR = Files.openFileReader(ApplicationSetup.COLLECTION_SPEC); String line = null; List<Path> paths = new ArrayList<Path>(); while ((line = specBR.readLine()) != null) { if (line.startsWith("#")) continue; paths.add(new Path(line)); } specBR.close(); FileInputFormat.setInputPaths(conf, paths.toArray(new Path[paths.size()])); // not sure if this is effective in YARN conf.setNumMapTasks(2000); // increase the heap usage conf.set("mapreduce.map.memory.mb", "6100"); conf.set("mapred.job.map.memory.mb", "6100"); conf.set("mapreduce.reduce.memory.mb", "6144"); conf.set("mapred.job.reduce.memory.mb", "6144"); conf.set("mapreduce.map.java.opts", "-Xmx6100m"); conf.set("mapred.map.child.java.opts", "-Xmx6100m"); conf.set("mapreduce.reduce.java.opts", "-Xmx6144m"); conf.set("mapred.reduce.child.opts", "-Xmx6144m"); //conf.setBoolean("mapred.used.genericoptionsparser", true) ; // This is the nasty thing in MapReduce v2 and YARN: They always prefer their ancient jars first. Set this on to say you don't like it conf.set("mapreduce.job.user.classpath.first", "true"); // increase the yarn memory to 10 GB conf.set("yarn.nodemanager.resource.memory-mb", "12288"); conf.set("yarn.nodemanager.resource.cpu-vcores", "16"); conf.set("yarn.scheduler.minimum-allocation-mb", "4096"); conf.setNumReduceTasks(numberOfReducers); if (numberOfReducers > 1) { if (docPartitioned) conf.setPartitionerClass(SplitEmittedTerm.SETPartitioner.class); else conf.setPartitionerClass(SplitEmittedTerm.SETPartitionerLowercaseAlphaTerm.class); } else { //for JUnit tests, we seem to need to restore the original partitioner class conf.setPartitionerClass(HashPartitioner.class); } /*JobID jobId = null; boolean ranOK = true; try{ RunningJob rj = JobClient.runJob(conf); jobId = rj.getID(); HadoopUtility.finishTerrierJob(conf); } catch (Exception e) { logger.error("Problem running job", e); e.printStackTrace(); ranOK = false; } if (jobId != null) { deleteTaskFiles(ApplicationSetup.TERRIER_INDEX_PATH, jobId); } */ //if (ranOK) //{ System.out.println("Merging indices"); if (!docPartitioned) { if (numberOfReducers > 1) mergeLexiconInvertedFiles(ApplicationSetup.TERRIER_INDEX_PATH, numberOfReducers); } Hadoop_BasicSinglePassIndexer.finish(ApplicationSetup.TERRIER_INDEX_PATH, docPartitioned ? numberOfReducers : 1, jf); //} System.out.println("Time Taken = " + ((System.currentTimeMillis() - time) / 1000) + " seconds"); jf.close(); return 0; }
From source file:de.tudarmstadt.lt.n2n.hadoop.GoogleSyntacticsJob.java
License:Apache License
@Override public int run(String[] args) throws Exception { JobConf conf = new JobConf(getConf(), GoogleSyntacticsJob.class); conf.setJobName(GoogleSyntacticsJob.class.getSimpleName()); conf.setMapperClass(GoogleSyntacticsJob3Mapper.class); conf.setReducerClass(GoogleSyntacticsJob3Reducer.class); conf.setInputFormat(TextInputFormat.class); conf.setOutputFormat(TextOutputFormat.class); conf.setOutputKeyClass(JoBimFormat.class); conf.setOutputValueClass(IntWritable.class); FileInputFormat.setInputPaths(conf, new Path(args[0])); FileOutputFormat.setOutputPath(conf, new Path(args[1])); // delete output path for testing purposes // FileSystem.get(conf).delete(new Path(args[1]), true); String extractorConfigurationFiles = conf.get(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS); if (extractorConfigurationFiles == null) { extractorConfigurationFiles = StringUtils.join(SHARED_CONSTANTS.DEFAULT_EXTRACTOR_CONFIGURATIONS, ','); System.out.format("Extractorconfigurationfile parameter not set. Assuming -D%s=%s %n", SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); conf.set(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); }/*from ww w . jav a 2s . c o m*/ String[] extractorConfigurationFilesArr = extractorConfigurationFiles.split(","); for (int i = 0; i < extractorConfigurationFilesArr.length; i++) DistributedCache.addFileToClassPath(new Path(extractorConfigurationFilesArr[i]), conf); JobClient.runJob(conf); return 0; }
From source file:de.tudarmstadt.lt.n2n.hadoop.GoogleSyntacticsJobDkbd.java
License:Apache License
@Override public void configure(JobConf job) { String extractorConfigurationFiles = job.get(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS); if (extractorConfigurationFiles == null) { extractorConfigurationFiles = StringUtils.join(SHARED_CONSTANTS.DEFAULT_EXTRACTOR_CONFIGURATIONS, ','); System.out.format("Extractorconfigurationfile parameter not set. Assuming -D%s=%s %n", SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); job.set(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); }//from w ww . j ava 2s. co m try { String[] extractorConfigurationFilesArr = extractorConfigurationFiles.split(","); for (int i = 0; i < extractorConfigurationFilesArr.length; i++) DistributedCache.addFileToClassPath(new Path(extractorConfigurationFilesArr[i]), job); } catch (IOException e) { e.printStackTrace(); } Text2CASInputFormat.setDocumentTextExtractorClass(job, KeyPlusValueAsDocumentExtractor.class); job.setOutputFormat(NullOutputFormat.class); // ignore the serialized cas and use only the output from the CasConsumer }
From source file:de.tudarmstadt.lt.n2n.hadoop.pipetests.GoogleSyntacticsJob2.java
License:Apache License
@Override public int run(String[] args) throws Exception { JobConf conf = new JobConf(getConf(), GoogleSyntacticsJob2.class); conf.setJobName(GoogleSyntacticsJob2.class.getSimpleName()); conf.setMapperClass(GoogleSyntacticsJob2Mapper.class); conf.setReducerClass(GoogleSyntacticsJob2Reducer.class); conf.setInputFormat(TextInputFormat.class); conf.setOutputFormat(TextOutputFormat.class); // conf.setMapOutputKeyClass(Text.class); // conf.setMapOutputValueClass(NullWritable.class); conf.setOutputKeyClass(JoBimFormat.class); conf.setOutputValueClass(IntWritable.class); args = new GenericOptionsParser(conf, args).getRemainingArgs(); FileInputFormat.setInputPaths(conf, new Path(args[0])); FileOutputFormat.setOutputPath(conf, new Path(args[1])); // delete output path for testing purposes // FileSystem.get(conf).delete(new Path(args[1]), true); String extractorConfigurationFiles = conf.get(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS); if (extractorConfigurationFiles == null) { extractorConfigurationFiles = StringUtils.join(SHARED_CONSTANTS.DEFAULT_EXTRACTOR_CONFIGURATIONS, ','); System.out.format("Extractorconfigurationfile parameter not set. Assuming -D%s=%s %n", SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); conf.set(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); }/* w w w .ja v a 2 s . c om*/ String[] extractorConfigurationFilesArr = extractorConfigurationFiles.split(","); for (int i = 0; i < extractorConfigurationFilesArr.length; i++) DistributedCache.addFileToClassPath(new Path(extractorConfigurationFilesArr[i]), conf); JobClient.runJob(conf); return 0; }
From source file:de.tudarmstadt.lt.n2n.hadoop.pipetests.GoogleSyntacticsJob4.java
License:Apache License
@Override public int run(String[] args) throws Exception { JobConf conf = new JobConf(getConf(), GoogleSyntacticsJob4.class); conf.setJobName(GoogleSyntacticsJob4.class.getSimpleName()); FileInputFormat.setInputPaths(conf, new Path(args[0])); FileOutputFormat.setOutputPath(conf, new Path(args[1])); // delete output path for testing purposes // FileSystem.get(conf).delete(new Path(args[1]), true); String extractorConfigurationFiles = conf.get(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS); if (extractorConfigurationFiles == null) { extractorConfigurationFiles = StringUtils.join(SHARED_CONSTANTS.DEFAULT_EXTRACTOR_CONFIGURATIONS, ','); System.out.format("Extractorconfigurationfile parameter not set. Assuming -D%s=%s %n", SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); conf.set(SHARED_CONSTANTS.PARAM_EXTRACTORCONFIGS, extractorConfigurationFiles); }/*w w w . j ava 2 s. c o m*/ String[] extractorConfigurationFilesArr = extractorConfigurationFiles.split(","); for (int i = 0; i < extractorConfigurationFilesArr.length; i++) DistributedCache.addFileToClassPath(new Path(extractorConfigurationFilesArr[i]), conf); conf.setMapperClass(GoogleSyntacticsJob4Mapper.class); conf.setInputFormat(TextInputFormat.class); conf.setMapOutputKeyClass(NullWritable.class); conf.setMapOutputValueClass(Text.class); conf.setNumReduceTasks(0); conf.setCombinerClass(IdentityReducer.class); JobClient.runJob(conf); return 0; }