List of usage examples for org.apache.hadoop.mapred JobConf setOutputKeyComparatorClass
public void setOutputKeyComparatorClass(Class<? extends RawComparator> theClass)
From source file:edu.stolaf.cs.wmrserver.HadoopEngine.java
License:Apache License
public void submit(JobRequest request, long submissionID, File mapperFile, File reducerFile, File packageDir, Path inputPath) throws ValidationException, NotFoundException, CompilationException, InternalException { // Generate job output path Path outputDir = new Path(_homeDir, "out"); Path outputPath;/*from w ww . j av a 2s.c om*/ try { FileSystem fs = outputDir.getFileSystem(new Configuration()); outputPath = JobServiceHandler.getNonexistantPath(outputDir, request.getName(), fs); } catch (IOException ex) { throw JobServiceHandler.wrapException("Could not construct output path.", ex); } JobConf conf = new JobConf(); conf.setJobName(request.getName()); // Set mapper and number of tasks if specified StreamJob.setStreamMapper(conf, mapperFile.toString()); if (request.isSetMapTasks()) conf.setNumMapTasks(request.getMapTasks()); // Set reducer and number of tasks if specified StreamJob.setStreamReducer(conf, reducerFile.toString()); if (request.isSetReduceTasks()) conf.setNumReduceTasks(request.getReduceTasks()); // Create and set job JAR, including necessary files ArrayList<String> jarFiles = new ArrayList<String>(); jarFiles.add(packageDir.toString()); String jarPath; try { jarPath = StreamJob.createJobJar(conf, jarFiles, _tempDir); } catch (IOException ex) { throw JobServiceHandler.wrapException("Could not create job jar.", ex); } if (jarPath != null) conf.setJar(jarPath); // TODO: This is a hack. Rewrite streaming to use DistributedCache. //conf.setPattern("mapreduce.job.jar.unpack.pattern", // Pattern.compile(".*")); // Set I/O formats and paths conf.setInputFormat(KeyValueTextInputFormat.class); conf.setOutputFormat(TextOutputFormat.class); conf.setOutputKeyClass(Text.class); conf.setOutputValueClass(Text.class); FileInputFormat.addInputPath(conf, inputPath); FileOutputFormat.setOutputPath(conf, outputPath); // Use numeric sort if appropriate conf.setBoolean(CONF_NUMERIC, request.isNumericSort()); if (request.isNumericSort()) { conf.setOutputKeyComparatorClass(KeyFieldBasedComparator.class); conf.setPartitionerClass(KeyFieldBasedPartitioner.class); conf.setKeyFieldComparatorOptions("-n"); conf.setKeyFieldPartitionerOptions("-n"); } // Set other job information conf.set(CONF_USER, request.getUser()); conf.set(CONF_LANGUAGE, request.getLanguage()); conf.set(CONF_MAPPER, request.getMapper()); conf.set(CONF_REDUCER, request.getReducer()); // Attempt to submit the job RunningJob job; try { JobClient client = new JobClient(new JobConf()); job = client.submitJob(conf); } catch (IOException ex) { throw JobServiceHandler.wrapException("There was a serious error while attempting to submit the job.", ex); } try { SubmissionDatabase.setSubmitted(submissionID); SubmissionDatabase.setHadoopID(submissionID, job.getID().toString()); } catch (SQLException ex) { throw JobServiceHandler.wrapException("Could not update submission in database.", ex); } }
From source file:mahout.analysis.MahoutOA.java
public static void main(String[] args) throws Exception { Path outPath = new Path(OUTPUT_PATH); JobConf conf = new JobConf(MahoutOA.class); conf.setJobName("Recommended books count"); conf.setMapperClass(MahoutOAMapper.class); conf.setMapOutputKeyClass(Text.class); conf.setMapOutputValueClass(IntWritable.class); conf.setReducerClass(MahoutOAReducer.class); FileInputFormat.setInputPaths(conf, new Path(args[0])); FileOutputFormat.setOutputPath(conf, outPath); try {//from ww w . j ava 2 s . c o m FileSystem dfs = FileSystem.get(outPath.toUri(), conf); if (dfs.exists(outPath)) { dfs.delete(outPath, true); } JobClient.runJob(conf); } catch (Exception e) { e.printStackTrace(); } JobConf conf1 = new JobConf(MahoutOA.class); Path outPathFinal = new Path(args[1]); conf1.setJobName("Recommended books count ordered"); conf1.setMapperClass(MahoutOASortMapper.class); conf1.setMapOutputKeyClass(IntWritable.class); conf1.setMapOutputValueClass(Text.class); conf1.setOutputKeyComparatorClass(MahoutOAComparator.class); conf1.setReducerClass(MahoutOASortReducer.class); FileInputFormat.setInputPaths(conf1, new Path(OUTPUT_PATH)); FileOutputFormat.setOutputPath(conf1, outPathFinal); try { FileSystem dfs = FileSystem.get(outPathFinal.toUri(), conf); if (dfs.exists(outPathFinal)) { dfs.delete(outPathFinal, true); } JobClient.runJob(conf1); } catch (Exception e) { e.printStackTrace(); } }
From source file:net.peacesoft.nutch.crawl.ReGenerator.java
License:Apache License
/** * Generate fetchlists in one or more segments. Whether to filter URLs or * not is read from the crawl.generate.filter property in the configuration * files. If the property is not found, the URLs are filtered. Same for the * normalisation./* w w w. j a va2 s . c om*/ * * @param dbDir Crawl database directory * @param segments Segments directory * @param numLists Number of reduce tasks * @param topN Number of top URLs to be selected * @param curTime Current time in milliseconds * * @return Path to generated segment or null if no entries were selected * * @throws IOException When an I/O error occurs */ public Path[] generate(Path dbDir, Path segments, int numLists, long topN, long curTime, boolean filter, boolean norm, boolean force, int maxNumSegments) throws IOException { try { Path tempDir = new Path( getConf().get("mapred.temp.dir", ".") + "/generate-temp-" + System.currentTimeMillis()); Path lock = new Path(dbDir, CrawlDb.LOCK_NAME); FileSystem fs = FileSystem.get(getConf()); LockUtil.createLockFile(fs, lock, force); SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss"); long start = System.currentTimeMillis(); LOG.info("ReGenerator: starting at " + sdf.format(start)); LOG.info("ReGenerator: Selecting best-scoring urls due for fetch."); LOG.info("ReGenerator: filtering: " + filter); LOG.info("ReGenerator: normalizing: " + norm); if (topN != Long.MAX_VALUE) { LOG.info("ReGenerator: topN: " + topN); } if ("true".equals(getConf().get(GENERATE_MAX_PER_HOST_BY_IP))) { LOG.info( "ReGenerator: GENERATE_MAX_PER_HOST_BY_IP will be ignored, use partition.url.mode instead"); } // map to inverted subset due for fetch, sort by score JobConf job = new NutchJob(getConf()); job.setJobName("generate: select from " + dbDir); if (numLists == -1) { // for politeness make numLists = job.getNumMapTasks(); // a partition per fetch task } if ("local".equals(job.get("mapred.job.tracker")) && numLists != 1) { // override LOG.info("ReGenerator: jobtracker is 'local', generating exactly one partition."); numLists = 1; } job.setLong(GENERATOR_CUR_TIME, curTime); // record real generation time long generateTime = System.currentTimeMillis(); job.setLong(Nutch.GENERATE_TIME_KEY, generateTime); job.setLong(GENERATOR_TOP_N, topN); job.setBoolean(GENERATOR_FILTER, filter); job.setBoolean(GENERATOR_NORMALISE, norm); job.setInt(GENERATOR_MAX_NUM_SEGMENTS, maxNumSegments); FileInputFormat.addInputPath(job, new Path(dbDir, CrawlDb.CURRENT_NAME)); job.setInputFormat(SequenceFileInputFormat.class); job.setMapperClass(Selector.class); job.setPartitionerClass(Selector.class); job.setReducerClass(Selector.class); FileOutputFormat.setOutputPath(job, tempDir); job.setOutputFormat(SequenceFileOutputFormat.class); job.setOutputKeyClass(FloatWritable.class); job.setOutputKeyComparatorClass(DecreasingFloatComparator.class); job.setOutputValueClass(SelectorEntry.class); job.setOutputFormat(GeneratorOutputFormat.class); try { JobClient.runJob(job); } catch (IOException e) { throw e; } // read the subdirectories generated in the temp // output and turn them into segments List<Path> generatedSegments = new ArrayList<Path>(); FileStatus[] status = fs.listStatus(tempDir); try { for (FileStatus stat : status) { Path subfetchlist = stat.getPath(); if (!subfetchlist.getName().startsWith("fetchlist-")) { continue; } // start a new partition job for this segment Path newSeg = partitionSegment(fs, segments, subfetchlist, numLists); generatedSegments.add(newSeg); } } catch (Exception e) { LOG.warn("ReGenerator: exception while partitioning segments, exiting ..."); fs.delete(tempDir, true); return null; } if (generatedSegments.size() == 0) { LOG.warn("ReGenerator: 0 records selected for fetching, exiting ..."); LockUtil.removeLockFile(fs, lock); fs.delete(tempDir, true); return null; } if (getConf().getBoolean(GENERATE_UPDATE_CRAWLDB, false)) { // update the db from tempDir Path tempDir2 = new Path( getConf().get("mapred.temp.dir", ".") + "/generate-temp-" + System.currentTimeMillis()); job = new NutchJob(getConf()); job.setJobName("generate: updatedb " + dbDir); job.setLong(Nutch.GENERATE_TIME_KEY, generateTime); for (Path segmpaths : generatedSegments) { Path subGenDir = new Path(segmpaths, CrawlDatum.GENERATE_DIR_NAME); FileInputFormat.addInputPath(job, subGenDir); } FileInputFormat.addInputPath(job, new Path(dbDir, CrawlDb.CURRENT_NAME)); job.setInputFormat(SequenceFileInputFormat.class); job.setMapperClass(CrawlDbUpdater.class); job.setReducerClass(CrawlDbUpdater.class); job.setOutputFormat(MapFileOutputFormat.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(CrawlDatum.class); FileOutputFormat.setOutputPath(job, tempDir2); try { JobClient.runJob(job); CrawlDb.install(job, dbDir); } catch (IOException e) { LockUtil.removeLockFile(fs, lock); fs.delete(tempDir, true); fs.delete(tempDir2, true); throw e; } fs.delete(tempDir2, true); } LockUtil.removeLockFile(fs, lock); fs.delete(tempDir, true); long end = System.currentTimeMillis(); LOG.info("ReGenerator: finished at " + sdf.format(end) + ", elapsed: " + TimingUtil.elapsedTime(start, end)); Path[] patharray = new Path[generatedSegments.size()]; return generatedSegments.toArray(patharray); } catch (Exception ex) { LOG.error("ReGenerator generate error: " + ex.toString(), ex); return null; } }
From source file:org.apache.avro.mapred.AvroJob.java
License:Apache License
private static void configureAvroShuffle(JobConf job) { job.setOutputKeyComparatorClass(AvroKeyComparator.class); job.setMapOutputKeyClass(AvroKey.class); job.setMapOutputValueClass(AvroValue.class); // add AvroSerialization to io.serializations Collection<String> serializations = job.getStringCollection("io.serializations"); if (!serializations.contains(AvroSerialization.class.getName())) { serializations.add(AvroSerialization.class.getName()); job.setStrings("io.serializations", serializations.toArray(new String[0])); }/* w w w .j a v a 2 s . co m*/ }
From source file:org.apache.avro.mapred.tether.TetherJob.java
License:Apache License
private static void setupTetherJob(JobConf job) throws IOException { job.setMapRunnerClass(TetherMapRunner.class); job.setPartitionerClass(TetherPartitioner.class); job.setReducerClass(TetherReducer.class); job.setInputFormat(TetherInputFormat.class); job.setOutputFormat(TetherOutputFormat.class); job.setOutputKeyClass(TetherData.class); job.setOutputKeyComparatorClass(TetherKeyComparator.class); job.setMapOutputValueClass(NullWritable.class); // set the map output key class to TetherData job.setMapOutputKeyClass(TetherData.class); // add TetherKeySerialization to io.serializations Collection<String> serializations = job.getStringCollection("io.serializations"); if (!serializations.contains(TetherKeySerialization.class.getName())) { serializations.add(TetherKeySerialization.class.getName()); job.setStrings("io.serializations", serializations.toArray(new String[0])); }//from w w w. ja va 2s. c o m // determine whether the executable should be added to the cache. if (job.getBoolean(TETHER_EXEC_CACHED, false)) { DistributedCache.addCacheFile(getExecutable(job), job); } }
From source file:org.apache.mahout.classifier.bayes.mapreduce.common.BayesFeatureDriver.java
License:Apache License
@Override public void runJob(Path input, Path output, BayesParameters params) throws IOException { Configurable client = new JobClient(); JobConf conf = new JobConf(BayesFeatureDriver.class); conf.setJobName("Bayes Feature Driver running over input: " + input); conf.setOutputKeyClass(StringTuple.class); conf.setOutputValueClass(DoubleWritable.class); conf.setPartitionerClass(FeaturePartitioner.class); conf.setOutputKeyComparatorClass(FeatureLabelComparator.class); FileInputFormat.setInputPaths(conf, input); FileOutputFormat.setOutputPath(conf, output); conf.setMapperClass(BayesFeatureMapper.class); conf.setInputFormat(KeyValueTextInputFormat.class); conf.setCombinerClass(BayesFeatureCombiner.class); conf.setReducerClass(BayesFeatureReducer.class); conf.setOutputFormat(BayesFeatureOutputFormat.class); conf.set("io.serializations", "org.apache.hadoop.io.serializer.JavaSerialization,org.apache.hadoop.io.serializer.WritableSerialization"); // this conf parameter needs to be set enable serialisation of conf values HadoopUtil.overwriteOutput(output);// w w w . j a v a 2s . c o m conf.set("bayes.parameters", params.toString()); client.setConf(conf); JobClient.runJob(conf); }
From source file:org.apache.nutch.crawl.Generator.java
License:Apache License
/** * Generate fetchlists in one or more segments. Whether to filter URLs or not * is read from the crawl.generate.filter property in the configuration files. * If the property is not found, the URLs are filtered. Same for the * normalisation.//from w w w . j ava2 s. c o m * * @param dbDir * Crawl database directory * @param segments * Segments directory * @param numLists * Number of reduce tasks * @param topN * Number of top URLs to be selected * @param curTime * Current time in milliseconds * * @return Path to generated segment or null if no entries were selected * * @throws IOException * When an I/O error occurs */ public Path[] generate(Path dbDir, Path segments, int numLists, long topN, long curTime, boolean filter, boolean norm, boolean force, int maxNumSegments) throws IOException { Path tempDir = new Path( getConf().get("mapred.temp.dir", ".") + "/generate-temp-" + System.currentTimeMillis()); Path lock = new Path(dbDir, CrawlDb.LOCK_NAME); FileSystem fs = FileSystem.get(getConf()); LockUtil.createLockFile(fs, lock, force); SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss"); long start = System.currentTimeMillis(); LOG.info("Generator: starting at " + sdf.format(start)); LOG.info("Generator: Selecting best-scoring urls due for fetch."); LOG.info("Generator: filtering: " + filter); LOG.info("Generator: normalizing: " + norm); if (topN != Long.MAX_VALUE) { LOG.info("Generator: topN: " + topN); } if ("true".equals(getConf().get(GENERATE_MAX_PER_HOST_BY_IP))) { LOG.info("Generator: GENERATE_MAX_PER_HOST_BY_IP will be ignored, use partition.url.mode instead"); } // map to inverted subset due for fetch, sort by score JobConf job = new NutchJob(getConf()); job.setJobName("generate: select from " + dbDir); if (numLists == -1) { // for politeness make numLists = job.getNumMapTasks(); // a partition per fetch task } if ("local".equals(job.get("mapred.job.tracker")) && numLists != 1) { // override LOG.info("Generator: jobtracker is 'local', generating exactly one partition."); numLists = 1; } job.setLong(GENERATOR_CUR_TIME, curTime); // record real generation time long generateTime = System.currentTimeMillis(); job.setLong(Nutch.GENERATE_TIME_KEY, generateTime); job.setLong(GENERATOR_TOP_N, topN); job.setBoolean(GENERATOR_FILTER, filter); job.setBoolean(GENERATOR_NORMALISE, norm); job.setInt(GENERATOR_MAX_NUM_SEGMENTS, maxNumSegments); FileInputFormat.addInputPath(job, new Path(dbDir, CrawlDb.CURRENT_NAME)); job.setInputFormat(SequenceFileInputFormat.class); job.setMapperClass(Selector.class); job.setPartitionerClass(Selector.class); job.setReducerClass(Selector.class); FileOutputFormat.setOutputPath(job, tempDir); job.setOutputFormat(SequenceFileOutputFormat.class); job.setOutputKeyClass(FloatWritable.class); job.setOutputKeyComparatorClass(DecreasingFloatComparator.class); job.setOutputValueClass(SelectorEntry.class); job.setOutputFormat(GeneratorOutputFormat.class); try { JobClient.runJob(job); } catch (IOException e) { throw e; } // read the subdirectories generated in the temp // output and turn them into segments List<Path> generatedSegments = new ArrayList<Path>(); FileStatus[] status = fs.listStatus(tempDir); try { for (FileStatus stat : status) { Path subfetchlist = stat.getPath(); if (!subfetchlist.getName().startsWith("fetchlist-")) continue; // start a new partition job for this segment Path newSeg = partitionSegment(fs, segments, subfetchlist, numLists); generatedSegments.add(newSeg); } } catch (Exception e) { LOG.warn("Generator: exception while partitioning segments, exiting ..."); fs.delete(tempDir, true); return null; } if (generatedSegments.size() == 0) { LOG.warn("Generator: 0 records selected for fetching, exiting ..."); LockUtil.removeLockFile(fs, lock); fs.delete(tempDir, true); return null; } if (getConf().getBoolean(GENERATE_UPDATE_CRAWLDB, false)) { // update the db from tempDir Path tempDir2 = new Path( getConf().get("mapred.temp.dir", ".") + "/generate-temp-" + System.currentTimeMillis()); job = new NutchJob(getConf()); job.setJobName("generate: updatedb " + dbDir); job.setLong(Nutch.GENERATE_TIME_KEY, generateTime); for (Path segmpaths : generatedSegments) { Path subGenDir = new Path(segmpaths, CrawlDatum.GENERATE_DIR_NAME); FileInputFormat.addInputPath(job, subGenDir); } FileInputFormat.addInputPath(job, new Path(dbDir, CrawlDb.CURRENT_NAME)); job.setInputFormat(SequenceFileInputFormat.class); job.setMapperClass(CrawlDbUpdater.class); job.setReducerClass(CrawlDbUpdater.class); job.setOutputFormat(MapFileOutputFormat.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(CrawlDatum.class); FileOutputFormat.setOutputPath(job, tempDir2); try { JobClient.runJob(job); CrawlDb.install(job, dbDir); } catch (IOException e) { LockUtil.removeLockFile(fs, lock); fs.delete(tempDir, true); fs.delete(tempDir2, true); throw e; } fs.delete(tempDir2, true); } LockUtil.removeLockFile(fs, lock); fs.delete(tempDir, true); long end = System.currentTimeMillis(); LOG.info("Generator: finished at " + sdf.format(end) + ", elapsed: " + TimingUtil.elapsedTime(start, end)); Path[] patharray = new Path[generatedSegments.size()]; return generatedSegments.toArray(patharray); }
From source file:org.apache.nutch.tools.FreeGenerator.java
License:Apache License
public int run(String[] args) throws Exception { if (args.length < 2) { System.err.println("Usage: FreeGenerator <inputDir> <segmentsDir> [-filter] [-normalize]"); System.err.println("\tinputDir\tinput directory containing one or more input files."); System.err.println("\t\tEach text file contains a list of URLs, one URL per line"); System.err.println("\tsegmentsDir\toutput directory, where new segment will be created"); System.err.println("\t-filter\trun current URLFilters on input URLs"); System.err.println("\t-normalize\trun current URLNormalizers on input URLs"); return -1; }/*from w w w.ja va2 s . c o m*/ boolean filter = false; boolean normalize = false; if (args.length > 2) { for (int i = 2; i < args.length; i++) { if (args[i].equals("-filter")) { filter = true; } else if (args[i].equals("-normalize")) { normalize = true; } else { LOG.error("Unknown argument: " + args[i] + ", exiting ..."); return -1; } } } SimpleDateFormat sdf = new SimpleDateFormat("yyyy-MM-dd HH:mm:ss"); long start = System.currentTimeMillis(); LOG.info("FreeGenerator: starting at " + sdf.format(start)); JobConf job = new NutchJob(getConf()); job.setBoolean(FILTER_KEY, filter); job.setBoolean(NORMALIZE_KEY, normalize); FileInputFormat.addInputPath(job, new Path(args[0])); job.setInputFormat(TextInputFormat.class); job.setMapperClass(FG.class); job.setMapOutputKeyClass(Text.class); job.setMapOutputValueClass(Generator.SelectorEntry.class); job.setPartitionerClass(URLPartitioner.class); job.setReducerClass(FG.class); String segName = Generator.generateSegmentName(); job.setNumReduceTasks(job.getNumMapTasks()); job.setOutputFormat(SequenceFileOutputFormat.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(CrawlDatum.class); job.setOutputKeyComparatorClass(Generator.HashComparator.class); FileOutputFormat.setOutputPath(job, new Path(args[1], new Path(segName, CrawlDatum.GENERATE_DIR_NAME))); try { JobClient.runJob(job); } catch (Exception e) { LOG.error("FAILED: " + StringUtils.stringifyException(e)); return -1; } long end = System.currentTimeMillis(); LOG.info("FreeGenerator: finished at " + sdf.format(end) + ", elapsed: " + TimingUtil.elapsedTime(start, end)); return 0; }
From source file:org.apache.pig.backend.hadoop.executionengine.mapreduceExec.MapReduceLauncher.java
License:Apache License
/** * Submit a Pig job to hadoop.//from ww w . j a v a2s . co m * * @param mapFuncs * a list of map functions to apply to the inputs. The cardinality of the list should * be the same as input's cardinality. * @param groupFuncs * a list of grouping functions to apply to the inputs. The cardinality of the list * should be the same as input's cardinality. * @param reduceFunc * the reduce function. * @param mapTasks * the number of map tasks to use. * @param reduceTasks * the number of reduce tasks to use. * @param input * a list of inputs * @param output * the path of the output. * @return an indicator of success or failure. * @throws IOException */ public boolean launchPig(POMapreduce pom) throws IOException { JobConf conf = new JobConf(config); setJobProperties(conf, pom); Properties properties = pom.pigContext.getProperties(); ConfigurationValidator.validatePigProperties(properties); String jobName = properties.getProperty(PigContext.JOB_NAME); conf.setJobName(jobName); boolean success = false; List<String> funcs = new ArrayList<String>(); if (pom.toMap != null) { for (EvalSpec es : pom.toMap) funcs.addAll(es.getFuncs()); } if (pom.groupFuncs != null) { for (EvalSpec es : pom.groupFuncs) funcs.addAll(es.getFuncs()); } if (pom.toReduce != null) { funcs.addAll(pom.toReduce.getFuncs()); } // create jobs.jar locally and pass it to hadoop File submitJarFile = File.createTempFile("Job", ".jar"); try { FileOutputStream fos = new FileOutputStream(submitJarFile); JarManager.createJar(fos, funcs, null, pom.pigContext); log.debug("Job jar size = " + submitJarFile.length()); conf.setJar(submitJarFile.getPath()); String user = System.getProperty("user.name"); conf.setUser(user != null ? user : "Pigster"); conf.set("pig.spill.size.threshold", properties.getProperty("pig.spill.size.threshold")); conf.set("pig.spill.gc.activation.size", properties.getProperty("pig.spill.gc.activation.size")); if (pom.reduceParallelism != -1) { conf.setNumReduceTasks(pom.reduceParallelism); } if (pom.toMap != null) { conf.set("pig.mapFuncs", ObjectSerializer.serialize(pom.toMap)); } if (pom.toCombine != null) { conf.set("pig.combineFunc", ObjectSerializer.serialize(pom.toCombine)); // this is to make sure that combiner is only called once // since we can't handle no combine or multiple combines conf.setCombineOnceOnly(true); } if (pom.groupFuncs != null) { conf.set("pig.groupFuncs", ObjectSerializer.serialize(pom.groupFuncs)); } if (pom.toReduce != null) { conf.set("pig.reduceFunc", ObjectSerializer.serialize(pom.toReduce)); } if (pom.toSplit != null) { conf.set("pig.splitSpec", ObjectSerializer.serialize(pom.toSplit)); } if (pom.pigContext != null) { conf.set("pig.pigContext", ObjectSerializer.serialize(pom.pigContext)); } conf.setMapRunnerClass(PigMapReduce.class); if (pom.toCombine != null) { conf.setCombinerClass(PigCombine.class); //conf.setCombinerClass(PigMapReduce.class); } if (pom.quantilesFile != null) { conf.set("pig.quantilesFile", pom.quantilesFile); } else { // this is not a sort job - can use byte comparison to speed up processing conf.setOutputKeyComparatorClass(PigWritableComparator.class); } if (pom.partitionFunction != null) { conf.setPartitionerClass(SortPartitioner.class); } conf.setReducerClass(PigMapReduce.class); conf.setInputFormat(PigInputFormat.class); conf.setOutputFormat(PigOutputFormat.class); // not used starting with 0.15 conf.setInputKeyClass(Text.class); // not used starting with 0.15 conf.setInputValueClass(Tuple.class); conf.setOutputKeyClass(Tuple.class); if (pom.userComparator != null) { conf.setOutputKeyComparatorClass(pom.userComparator); } conf.setOutputValueClass(IndexedTuple.class); conf.set("pig.inputs", ObjectSerializer.serialize(pom.inputFileSpecs)); conf.setOutputPath(new Path(pom.outputFileSpec.getFileName())); conf.set("pig.storeFunc", ObjectSerializer.serialize(pom.outputFileSpec.getFuncSpec())); // Setup the DistributedCache for this job setupDistributedCache(pom.pigContext, conf, pom.properties, "pig.streaming.ship.files", true); setupDistributedCache(pom.pigContext, conf, pom.properties, "pig.streaming.cache.files", false); // Setup the logs directory for this job String jobOutputFileName = pom.pigContext.getJobOutputFile(); if (jobOutputFileName != null && jobOutputFileName.length() > 0) { Path jobOutputFile = new Path(pom.pigContext.getJobOutputFile()); conf.set("pig.output.dir", jobOutputFile.getParent().toString()); conf.set("pig.streaming.log.dir", new Path(jobOutputFile, LOG_DIR).toString()); } // // Now, actually submit the job (using the submit name) // JobClient jobClient = execEngine.getJobClient(); RunningJob status = jobClient.submitJob(conf); log.debug("submitted job: " + status.getJobID()); long sleepTime = 1000; double lastQueryProgress = -1.0; int lastJobsQueued = -1; double lastMapProgress = -1.0; double lastReduceProgress = -1.0; while (true) { try { Thread.sleep(sleepTime); } catch (Exception e) { } if (status.isComplete()) { success = status.isSuccessful(); if (log.isDebugEnabled()) { StringBuilder sb = new StringBuilder(); sb.append("Job finished "); sb.append((success ? "" : "un")); sb.append("successfully"); log.debug(sb.toString()); } if (success) { mrJobNumber++; } double queryProgress = ((double) mrJobNumber) / ((double) numMRJobs); if (queryProgress > lastQueryProgress) { if (log.isInfoEnabled()) { StringBuilder sbProgress = new StringBuilder(); sbProgress.append("Pig progress = "); sbProgress.append(((int) (queryProgress * 100))); sbProgress.append("%"); log.info(sbProgress.toString()); } lastQueryProgress = queryProgress; } break; } else // still running { double mapProgress = status.mapProgress(); double reduceProgress = status.reduceProgress(); if (lastMapProgress != mapProgress || lastReduceProgress != reduceProgress) { if (log.isDebugEnabled()) { StringBuilder sbProgress = new StringBuilder(); sbProgress.append("Hadoop job progress: Map="); sbProgress.append((int) (mapProgress * 100)); sbProgress.append("% Reduce="); sbProgress.append((int) (reduceProgress * 100)); sbProgress.append("%"); log.debug(sbProgress.toString()); } lastMapProgress = mapProgress; lastReduceProgress = reduceProgress; } double numJobsCompleted = mrJobNumber; double thisJobProgress = (mapProgress + reduceProgress) / 2.0; double queryProgress = (numJobsCompleted + thisJobProgress) / ((double) numMRJobs); if (queryProgress > lastQueryProgress) { if (log.isInfoEnabled()) { StringBuilder sbProgress = new StringBuilder(); sbProgress.append("Pig progress = "); sbProgress.append(((int) (queryProgress * 100))); sbProgress.append("%"); log.info(sbProgress.toString()); } lastQueryProgress = queryProgress; } } } // bug 1030028: if the input file is empty; hadoop doesn't create the output file! Path outputFile = conf.getOutputPath(); String outputName = outputFile.getName(); int colon = outputName.indexOf(':'); if (colon != -1) { outputFile = new Path(outputFile.getParent(), outputName.substring(0, colon)); } try { ElementDescriptor descriptor = ((HDataStorage) (pom.pigContext.getDfs())) .asElement(outputFile.toString()); if (success && !descriptor.exists()) { // create an empty output file PigFile f = new PigFile(outputFile.toString(), false); f.store(BagFactory.getInstance().newDefaultBag(), new PigStorage(), pom.pigContext); } } catch (DataStorageException e) { throw WrappedIOException.wrap("Failed to obtain descriptor for " + outputFile.toString(), e); } if (!success) { // go find the error messages getErrorMessages(jobClient.getMapTaskReports(status.getJobID()), "map"); getErrorMessages(jobClient.getReduceTaskReports(status.getJobID()), "reduce"); } else { long timeSpent = 0; // NOTE: this call is crashing due to a bug in Hadoop; the bug is known and the patch has not been applied yet. TaskReport[] mapReports = jobClient.getMapTaskReports(status.getJobID()); TaskReport[] reduceReports = jobClient.getReduceTaskReports(status.getJobID()); for (TaskReport r : mapReports) { timeSpent += (r.getFinishTime() - r.getStartTime()); } for (TaskReport r : reduceReports) { timeSpent += (r.getFinishTime() - r.getStartTime()); } totalHadoopTimeSpent += timeSpent; } } catch (Exception e) { // Do we need different handling for different exceptions e.printStackTrace(); throw WrappedIOException.wrap(e); } finally { submitJarFile.delete(); } return success; }
From source file:org.apache.sysml.runtime.controlprogram.parfor.ResultMergeRemoteMR.java
License:Apache License
@SuppressWarnings({ "unused", "deprecation" })
protected void executeMerge(String fname, String fnameNew, String[] srcFnames, InputInfo ii, OutputInfo oi,
long rlen, long clen, int brlen, int bclen) throws DMLRuntimeException {
String jobname = "ParFor-RMMR";
long t0 = DMLScript.STATISTICS ? System.nanoTime() : 0;
JobConf job = new JobConf(ResultMergeRemoteMR.class);
job.setJobName(jobname + _pfid);/* ww w.ja va 2 s .c om*/
//maintain dml script counters
Statistics.incrementNoOfCompiledMRJobs();
//warning for textcell/binarycell without compare
boolean withCompare = (fname != null);
if ((oi == OutputInfo.TextCellOutputInfo || oi == OutputInfo.BinaryCellOutputInfo) && !withCompare
&& ResultMergeLocalFile.ALLOW_COPY_CELLFILES)
LOG.warn("Result merge for " + OutputInfo.outputInfoToString(oi)
+ " without compare can be realized more efficiently with LOCAL_FILE than REMOTE_MR.");
try {
Path pathCompare = null;
Path pathNew = new Path(fnameNew);
/////
//configure the MR job
if (withCompare) {
FileSystem fs = IOUtilFunctions.getFileSystem(pathNew, job);
pathCompare = new Path(fname).makeQualified(fs);
MRJobConfiguration.setResultMergeInfo(job, pathCompare.toString(), ii,
LocalFileUtils.getWorkingDir(LocalFileUtils.CATEGORY_RESULTMERGE), rlen, clen, brlen,
bclen);
} else
MRJobConfiguration.setResultMergeInfo(job, "null", ii,
LocalFileUtils.getWorkingDir(LocalFileUtils.CATEGORY_RESULTMERGE), rlen, clen, bclen,
bclen);
//set mappers, reducers, combiners
job.setMapperClass(ResultMergeRemoteMapper.class);
job.setReducerClass(ResultMergeRemoteReducer.class);
if (oi == OutputInfo.TextCellOutputInfo) {
job.setMapOutputKeyClass(MatrixIndexes.class);
job.setMapOutputValueClass(TaggedMatrixCell.class);
job.setOutputKeyClass(NullWritable.class);
job.setOutputValueClass(Text.class);
} else if (oi == OutputInfo.BinaryCellOutputInfo) {
job.setMapOutputKeyClass(MatrixIndexes.class);
job.setMapOutputValueClass(TaggedMatrixCell.class);
job.setOutputKeyClass(MatrixIndexes.class);
job.setOutputValueClass(MatrixCell.class);
} else if (oi == OutputInfo.BinaryBlockOutputInfo) {
//setup partitioning, grouping, sorting for composite key (old API)
job.setPartitionerClass(ResultMergeRemotePartitioning.class); //partitioning
job.setOutputValueGroupingComparator(ResultMergeRemoteGrouping.class); //grouping
job.setOutputKeyComparatorClass(ResultMergeRemoteSorting.class); //sorting
job.setMapOutputKeyClass(ResultMergeTaggedMatrixIndexes.class);
job.setMapOutputValueClass(TaggedMatrixBlock.class);
job.setOutputKeyClass(MatrixIndexes.class);
job.setOutputValueClass(MatrixBlock.class);
}
//set input format
job.setInputFormat(ii.inputFormatClass);
//set the input path
Path[] paths = null;
if (withCompare) {
paths = new Path[srcFnames.length + 1];
paths[0] = pathCompare;
for (int i = 1; i < paths.length; i++)
paths[i] = new Path(srcFnames[i - 1]);
} else {
paths = new Path[srcFnames.length];
for (int i = 0; i < paths.length; i++)
paths[i] = new Path(srcFnames[i]);
}
FileInputFormat.setInputPaths(job, paths);
//set output format
job.setOutputFormat(oi.outputFormatClass);
//set output path
MapReduceTool.deleteFileIfExistOnHDFS(fnameNew);
FileOutputFormat.setOutputPath(job, pathNew);
//////
//set optimization parameters
//set the number of mappers and reducers
//job.setNumMapTasks( _numMappers ); //use default num mappers
long reducerGroups = _numReducers;
if (oi == OutputInfo.BinaryBlockOutputInfo)
reducerGroups = Math.max(rlen / brlen, 1) * Math.max(clen / bclen, 1);
else //textcell/binarycell
reducerGroups = Math.max((rlen * clen) / StagingFileUtils.CELL_BUFFER_SIZE, 1);
job.setNumReduceTasks((int) Math.min(_numReducers, reducerGroups));
//disable automatic tasks timeouts and speculative task exec
job.setInt(MRConfigurationNames.MR_TASK_TIMEOUT, 0);
job.setMapSpeculativeExecution(false);
//set up preferred custom serialization framework for binary block format
if (MRJobConfiguration.USE_BINARYBLOCK_SERIALIZATION)
MRJobConfiguration.addBinaryBlockSerializationFramework(job);
//set up custom map/reduce configurations
DMLConfig config = ConfigurationManager.getDMLConfig();
MRJobConfiguration.setupCustomMRConfigurations(job, config);
//enables the reuse of JVMs (multiple tasks per MR task)
if (_jvmReuse)
job.setNumTasksToExecutePerJvm(-1); //unlimited
//enables compression - not conclusive for different codecs (empirically good compression ratio, but significantly slower)
//job.set(MRConfigurationNames.MR_MAP_OUTPUT_COMPRESS, "true");
//job.set(MRConfigurationNames.MR_MAP_OUTPUT_COMPRESS_CODEC, "org.apache.hadoop.io.compress.GzipCodec");
//set the replication factor for the results
job.setInt(MRConfigurationNames.DFS_REPLICATION, _replication);
//set the max number of retries per map task
// disabled job-level configuration to respect cluster configuration
// note: this refers to hadoop2, hence it never had effect on mr1
//job.setInt(MRConfigurationNames.MR_MAP_MAXATTEMPTS, _max_retry);
//set unique working dir
MRJobConfiguration.setUniqueWorkingDir(job);
/////
// execute the MR job
JobClient.runJob(job);
//maintain dml script counters
Statistics.incrementNoOfExecutedMRJobs();
} catch (Exception ex) {
throw new DMLRuntimeException(ex);
}
if (DMLScript.STATISTICS) {
long t1 = System.nanoTime();
Statistics.maintainCPHeavyHitters("MR-Job_" + jobname, t1 - t0);
}
}