List of usage examples for org.apache.hadoop.mapred JobConf get
public String get(String name)
name property, null if no such property exists. From source file:com.yolodata.tbana.hadoop.mapred.csv.CSVTextInputFormat.java
License:Open Source License
@Override public RecordReader<LongWritable, List<Text>> getRecordReader(InputSplit split, JobConf jobConf, Reporter reporter) throws IOException { if (jobConf.get(CSVLineRecordReader.FORMAT_DELIMITER) == null || jobConf.get(CSVLineRecordReader.FORMAT_SEPARATOR) == null) { throw new IOException("CSVTextInputFormat: missing parameter delimiter/separator"); }//from w w w . ja v a 2 s . c o m CSVLineRecordReader reader = new CSVLineRecordReader(); reader.initialize(split, jobConf); return reader; }
From source file:com.yolodata.tbana.hadoop.mapred.splunk.split.SplitProvider.java
License:Open Source License
protected int getNumberOfSplits(JobConf conf, int defaultValue) { if (conf.get(SplunkInputFormat.INPUTFORMAT_SPLITS) != null) defaultValue = Integer.parseInt(conf.get(SplunkInputFormat.INPUTFORMAT_SPLITS)); return defaultValue; }
From source file:com.yolodata.tbana.hadoop.mapred.splunk.SplunkDataQueryFactory.java
License:Open Source License
public static SplunkDataQuery createWithJobConf(JobConf jobConf) { String earliest = jobConf.get(ShuttlInputFormatConstants.EARLIEST_TIME); String latest = jobConf.get(ShuttlInputFormatConstants.LATEST_TIME); String indexes = jobConf.get(ShuttlInputFormatConstants.INDEX_LIST); String[] indexList = indexes.split(SplunkDataQuery.INDEX_LIST_SEPARATOR); return new SplunkDataQuery(DateTime.parse(earliest), DateTime.parse(latest), indexList); }
From source file:com.zjy.mongo.mapred.output.MongoRecordWriter.java
License:Apache License
/** * Create a new MongoRecordWriter.// w w w. jav a 2 s . c om * @param conf the job configuration */ public MongoRecordWriter(final JobConf conf) { super(Collections.<DBCollection>emptyList(), new TaskAttemptContextImpl(conf, TaskAttemptID.forName(conf.get("mapred.task.id")))); configuration = conf; }
From source file:crunch.MaxTemperature.java
License:Apache License
@Test
public void inputPathProperty() throws Exception {
JobConf conf = new JobConf();
FileInputFormat.setInputPaths(conf, new Path("/{a,b}"), new Path("/{c,d}"));
assertThat(conf.get("mapred.input.dir"), is("file:/{a\\,b},file:/{c\\,d}"));
}// w w w . j av a2 s .c o m
From source file:DataModifier.UserSongMapper.java
public void configure(JobConf job) { user = job.get("userID"); }
From source file:de.l3s.streamcorpus.mapreduce.TerrierIndexing.java
License:Mozilla Public License
/** Starts the MapReduce indexing. * @param args/*from w ww . j a 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/*w w w . ja va2 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(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.FilterByVocabularyJob.java
License:Apache License
@Override public int run(String[] args) throws Exception { JobConf conf = new JobConf(getConf(), FilterByVocabularyJob.class); conf.setJobName(FilterByVocabularyJob.class.getSimpleName()); conf.setMapperClass(FilterByVocabularyMapper.class); conf.setReducerClass(IdentityReducer.class); conf.setInputFormat(KeyValueTextInputFormat.class); conf.setOutputFormat(TextOutputFormat.class); conf.setOutputKeyClass(Text.class); conf.setOutputValueClass(Text.class); String word_list_file = conf.get(SHARED_CONSTANTS.PARAM_WORD_LIST); if (word_list_file == null) throw new MissingArgumentException( "Please specify word list with '-Dnlkg.filterbywordsfile=<path-to-file-in-hdfs>'."); DistributedCache.addFileToClassPath(new Path(word_list_file), conf); FileInputFormat.setInputPaths(conf, new Path(args[0])); FileOutputFormat.setOutputPath(conf, new Path(args[1])); JobClient.runJob(conf);/*from w w w .ja v a2 s . c om*/ 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 w w w. j a v a2s . co 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; }