List of usage examples for org.apache.hadoop.mapred JobConf setInt
public void setInt(String name, int value)
name property to an int. From source file:cascading.flow.FlowStep.java
License:Open Source License
protected JobConf getJobConf(JobConf parentConf) throws IOException { JobConf conf = parentConf == null ? new JobConf() : new JobConf(parentConf); // set values first so they can't break things downstream if (hasProperties()) { for (Map.Entry entry : getProperties().entrySet()) conf.set(entry.getKey().toString(), entry.getValue().toString()); }//from w w w .j a va2 s . co m // disable warning conf.setBoolean("mapred.used.genericoptionsparser", true); conf.setJobName(getStepName()); conf.setOutputKeyClass(Tuple.class); conf.setOutputValueClass(Tuple.class); conf.setMapperClass(FlowMapper.class); conf.setReducerClass(FlowReducer.class); // set for use by the shuffling phase TupleSerialization.setSerializations(conf); initFromSources(conf); initFromSink(conf); initFromTraps(conf); if (sink.getScheme().getNumSinkParts() != 0) { // if no reducer, set num map tasks to control parts if (getGroup() != null) conf.setNumReduceTasks(sink.getScheme().getNumSinkParts()); else conf.setNumMapTasks(sink.getScheme().getNumSinkParts()); } conf.setOutputKeyComparatorClass(TupleComparator.class); if (getGroup() == null) { conf.setNumReduceTasks(0); // disable reducers } else { // must set map output defaults when performing a reduce conf.setMapOutputKeyClass(Tuple.class); conf.setMapOutputValueClass(Tuple.class); // handles the case the groupby sort should be reversed if (getGroup().isSortReversed()) conf.setOutputKeyComparatorClass(ReverseTupleComparator.class); addComparators(conf, "cascading.group.comparator", getGroup().getGroupingSelectors()); if (getGroup().isGroupBy()) addComparators(conf, "cascading.sort.comparator", getGroup().getSortingSelectors()); if (!getGroup().isGroupBy()) { conf.setPartitionerClass(CoGroupingPartitioner.class); conf.setMapOutputKeyClass(IndexTuple.class); // allows groups to be sorted by index conf.setMapOutputValueClass(IndexTuple.class); conf.setOutputKeyComparatorClass(IndexTupleCoGroupingComparator.class); // sorts by group, then by index conf.setOutputValueGroupingComparator(CoGroupingComparator.class); } if (getGroup().isSorted()) { conf.setPartitionerClass(GroupingPartitioner.class); conf.setMapOutputKeyClass(TuplePair.class); if (getGroup().isSortReversed()) conf.setOutputKeyComparatorClass(ReverseGroupingSortingComparator.class); else conf.setOutputKeyComparatorClass(GroupingSortingComparator.class); // no need to supply a reverse comparator, only equality is checked conf.setOutputValueGroupingComparator(GroupingComparator.class); } } // perform last so init above will pass to tasks conf.setInt("cascading.flow.step.id", id); conf.set("cascading.flow.step", Util.serializeBase64(this)); return conf; }
From source file:cascading.flow.FlowStep.java
License:Open Source License
private void addComparators(JobConf conf, String property, Map<String, Fields> map) throws IOException { Iterator<Fields> fieldsIterator = map.values().iterator(); if (!fieldsIterator.hasNext()) return;//from w w w .j av a2s . c om Fields fields = fieldsIterator.next(); if (fields.hasComparators()) { conf.set(property, Util.serializeBase64(fields)); return; } // use resolved fields if there are no comparators. Set<Scope> previousScopes = getPreviousScopes(getGroup()); fields = previousScopes.iterator().next().getOutValuesFields(); if (fields.size() != 0) // allows fields.UNKNOWN to be used conf.setInt(property + ".size", fields.size()); return; }
From source file:com.benchmark.mapred.SleepJob.java
License:Apache License
public JobConf setupJobConf(int numMapper, int numReducer, long mapSleepTime, int mapSleepCount, long reduceSleepTime, int reduceSleepCount) { JobConf job = new JobConf(getConf(), SleepJob.class); job.setNumMapTasks(numMapper);/*from w ww.ja va 2s.com*/ job.setNumReduceTasks(numReducer); job.setMapperClass(SleepJob.class); job.setMapOutputKeyClass(IntWritable.class); job.setMapOutputValueClass(NullWritable.class); job.setReducerClass(SleepJob.class); job.setOutputFormat(NullOutputFormat.class); job.setInputFormat(SleepInputFormat.class); job.setPartitionerClass(SleepJob.class); job.setSpeculativeExecution(false); job.setJobName("Sleep job"); FileInputFormat.addInputPath(job, new Path("ignored")); job.setLong("sleep.job.map.sleep.time", mapSleepTime); job.setLong("sleep.job.reduce.sleep.time", reduceSleepTime); job.setInt("sleep.job.map.sleep.count", mapSleepCount); job.setInt("sleep.job.reduce.sleep.count", reduceSleepCount); return job; }
From source file:com.benchmark.mapred.terasort.TeraSort.java
License:Apache License
public int run(String[] args) throws Exception { LOG.info("starting"); JobConf job = (JobConf) getConf(); Path inputDir = new Path(args[0]); if (args.length != 3) { System.out.println("ERROR: Wrong number of parameters: " + args.length + " instead of 3."); }/* w w w.j av a 2 s . c o m*/ inputDir = inputDir.makeQualified(inputDir.getFileSystem(job)); Path partitionFile = new Path(inputDir, TeraInputFormat.PARTITION_FILENAME); URI partitionUri = new URI(partitionFile.toString() + "#" + TeraInputFormat.PARTITION_FILENAME); TeraInputFormat.setInputPaths(job, new Path(args[0])); FileOutputFormat.setOutputPath(job, new Path(args[1])); job.setJobName("TeraSort"); job.setJarByClass(TeraSort.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(Text.class); job.setInputFormat(TeraInputFormat.class); job.setOutputFormat(TeraOutputFormat.class); job.setPartitionerClass(TotalOrderPartitioner.class); job.setNumReduceTasks(Integer.parseInt(args[2])); TeraInputFormat.writePartitionFile(job, partitionFile); DistributedCache.addCacheFile(partitionUri, job); DistributedCache.createSymlink(job); job.setInt("dfs.replication", 1); TeraOutputFormat.setFinalSync(job, true); Date startIteration = new Date(); JobClient.runJob(job); Date endIteration = new Date(); System.out.println( "The iteration took " + (endIteration.getTime() - startIteration.getTime()) / 1000 + " seconds."); LOG.info("done"); return 0; }
From source file:com.cloudera.circus.test.TestXTest.java
License:Open Source License
@Test @TestHadoop// w w w.ja v a 2s . co m public void testHadoopMapReduce() throws Exception { JobConf conf = getHadoopConf(); FileSystem fs = FileSystem.get(conf); JobClient jobClient = new JobClient(conf); try { Path inputDir = new Path(getHadoopTestDir(), "input"); Path outputDir = new Path(getHadoopTestDir(), "output"); fs.mkdirs(inputDir); Writer writer = new OutputStreamWriter(fs.create(new Path(inputDir, "data.txt"))); writer.write("a\n"); writer.write("b\n"); writer.write("c\n"); writer.close(); JobConf jobConf = getHadoopConf(); jobConf.setInt("mapred.map.tasks", 1); jobConf.setInt("mapred.map.max.attempts", 1); jobConf.setInt("mapred.reduce.max.attempts", 1); jobConf.set("mapred.input.dir", inputDir.toString()); jobConf.set("mapred.output.dir", outputDir.toString()); final RunningJob runningJob = jobClient.submitJob(jobConf); waitFor(60 * 1000, true, new Predicate() { @Override public boolean evaluate() throws Exception { return runningJob.isComplete(); } }); Assert.assertTrue(runningJob.isSuccessful()); Assert.assertTrue(fs.exists(new Path(outputDir, "part-00000"))); BufferedReader reader = new BufferedReader( new InputStreamReader(fs.open(new Path(outputDir, "part-00000")))); Assert.assertTrue(reader.readLine().trim().endsWith("a")); Assert.assertTrue(reader.readLine().trim().endsWith("b")); Assert.assertTrue(reader.readLine().trim().endsWith("c")); Assert.assertNull(reader.readLine()); reader.close(); } finally { fs.close(); jobClient.close(); } }
From source file:com.cloudera.recordservice.avro.mapred.ColorCount.java
License:Apache License
/** * Run the MR1 color count with generic records, and return a map of favorite colors to * the number of users.//from w w w.ja v a 2 s . c o m */ public static java.util.Map<String, Integer> countColors() throws IOException { String output = TestUtil.getTempDirectory(); Path outputPath = new Path(output); JobConf conf = new JobConf(ColorCount.class); conf.setJobName("MR1 Color Count With Generic Records"); conf.setInt("mapreduce.job.reduces", 1); conf.setBoolean(com.cloudera.recordservice.avro.AvroJob.USE_RECORD_SERVICE_INPUT_FORMAT_CONF_KEY, true); com.cloudera.recordservice.avro.AvroJob.setInputFormat(conf, org.apache.avro.mapred.AvroInputFormat.class); RecordServiceConfig.setInputTable(conf, "rs", "users"); FileOutputFormat.setOutputPath(conf, outputPath); AvroJob.setMapperClass(conf, Map.class); AvroJob.setReducerClass(conf, Reduce.class); AvroJob.setOutputSchema(conf, Pair.getPairSchema(Schema.create(Schema.Type.STRING), Schema.create(Schema.Type.INT))); JobClient.runJob(conf); // Read the result and return it. Since we set the number of reducers to 1, // there is always just one file containing the value. SeekableInput input = new FsInput(new Path(output + "/part-00000.avro"), conf); DatumReader<GenericRecord> reader = new GenericDatumReader<GenericRecord>(); FileReader<GenericRecord> fileReader = DataFileReader.openReader(input, reader); java.util.Map<String, Integer> colorMap = new HashMap<String, Integer>(); for (GenericRecord datum : fileReader) { colorMap.put(datum.get(0).toString(), Integer.parseInt(datum.get(1).toString())); } return colorMap; }
From source file:com.cloudera.recordservice.avro.mapreduce.ColorCount.java
License:Apache License
/** * Run the MR2 color count with generic records, and return a map of favorite colors to * the number of users.//from w w w. j ava 2 s .c o m */ public static java.util.Map<String, Integer> countColors() throws IOException, ClassNotFoundException, InterruptedException { String output = TestUtil.getTempDirectory(); Path outputPath = new Path(output); JobConf conf = new JobConf(ColorCount.class); conf.setInt("mapreduce.job.reduces", 1); Job job = Job.getInstance(conf); job.setJarByClass(ColorCount.class); job.setJobName("MR2 Color Count With Generic Records"); RecordServiceConfig.setInputTable(job.getConfiguration(), "rs", "users"); job.setInputFormatClass(com.cloudera.recordservice.avro.mapreduce.AvroKeyInputFormat.class); FileOutputFormat.setOutputPath(job, outputPath); job.setMapperClass(Map.class); job.setMapOutputKeyClass(Text.class); job.setMapOutputValueClass(IntWritable.class); job.setOutputFormatClass(AvroKeyValueOutputFormat.class); job.setReducerClass(Reduce.class); AvroJob.setOutputKeySchema(job, Schema.create(Schema.Type.STRING)); AvroJob.setOutputValueSchema(job, Schema.create(Schema.Type.INT)); job.waitForCompletion(false); // Read the result and return it. Since we set the number of reducers to 1, // there is always just one file containing the value. SeekableInput input = new FsInput(new Path(output + "/part-r-00000.avro"), conf); DatumReader<GenericRecord> reader = new GenericDatumReader<GenericRecord>(); FileReader<GenericRecord> fileReader = DataFileReader.openReader(input, reader); java.util.Map<String, Integer> colorMap = new HashMap<String, Integer>(); for (GenericRecord datum : fileReader) { colorMap.put(datum.get(0).toString(), Integer.parseInt(datum.get(1).toString())); } return colorMap; }
From source file:com.cloudera.recordservice.mapreduce.testapps.RecordCount.java
License:Apache License
public static long countRecords(String path) throws IOException { String output = TestUtil.getTempDirectory(); Path inputPath = new Path(path); Path outputPath = new Path(output); JobConf conf = new JobConf(RecordCount.class); conf.setJobName("recordcount"); conf.setOutputKeyClass(NullWritable.class); conf.setOutputValueClass(LongWritable.class); conf.setInt("mapreduce.job.reduces", 1); conf.setMapperClass(Map.class); conf.setCombinerClass(Reduce.class); conf.setReducerClass(Reduce.class); conf.setInputFormat(com.cloudera.recordservice.mapred.TextInputFormat.class); conf.setOutputFormat(TextOutputFormat.class); FileInputFormat.setInputPaths(conf, inputPath); FileOutputFormat.setOutputPath(conf, outputPath); JobClient.runJob(conf);/*from w ww . j av a 2 s . c o m*/ // Read the result and return it. Since we set the number of reducers to 1, // there is always just one file containing the value. FileSystem fs = outputPath.getFileSystem(conf); FSDataInputStream resultStream = fs.open(new Path(output + "/part-00000")); byte[] bytes = new byte[16]; int length = resultStream.read(bytes); String result = new String(bytes, 0, length).trim(); return Long.parseLong(result); }
From source file:com.finderbots.miner.MinerWorkflow.java
License:Apache License
public static Flow createWebMiningWorkflow(Path crawlDbPath, Path curLoopDirPath, FetcherPolicy fetcherPolicy, UserAgent userAgent, MinerOptions options, BaseUrlFilter crawlUrlFilter, BaseUrlFilter mineUrlFilter) throws IOException, InterruptedException { // Fetch at most 200 pages, max size of 128K, complete mode, from the current dir. // HTML only. // We want to extract the cleaned up HTML, and pass that to the parser, which will // be specified via options.getAnalyzer. From this we'll get outlinks, page score, and // any results. JobConf conf = HadoopUtils.getDefaultJobConf(CrawlConfig.CRAWL_STACKSIZE_KB); boolean isLocal = HadoopUtils.isJobLocal(conf); int numReducers = HadoopUtils.getNumReducers(conf); conf.setNumReduceTasks(numReducers); conf.setInt("mapred.min.split.size", 64 * 1024 * 1024); Properties props = HadoopUtils.getDefaultProperties(MinerWorkflow.class, false, conf); FileSystem fs = crawlDbPath.getFileSystem(conf); // Input : the crawldb if (!fs.exists(crawlDbPath)) { throw new RuntimeException("CrawlDb not found"); }//w w w . j a v a 2s. c o m //Tap inputSource = new Hfs(new TextDelimited(CrawlDbDatum.FIELDS, "\t", CrawlDbDatum.TYPES), crawlDbPath.toString()); Tap inputSource = new Hfs(new SequenceFile(CrawlDbDatum.FIELDS), crawlDbPath.toString(), true); Pipe importPipe = new Pipe("import pipe"); // Split into tuples that are to be fetched and that have already been fetched SplitterAssembly splitter = new SplitterAssembly(importPipe, new SplitFetchedUnfetchedSSCrawlDatums()); Pipe finishedDatumsFromDb = new Pipe("finished datums from db", splitter.getRHSPipe()); Pipe urlsToFetchPipe = splitter.getLHSPipe(); // Limit to MAX_DISTRIBUTED_FETCH if running in real cluster, // or MAX_LOCAL_FETCH if running locally. So first we sort the entries // from high to low by links score. // TODO add unit test urlsToFetchPipe = new GroupBy(urlsToFetchPipe, new Fields(CrawlDbDatum.LINKS_SCORE_FIELD), true); long maxToFetch = HadoopUtils.isJobLocal(conf) ? MAX_LOCAL_FETCH : MAX_DISTRIBUTED_FETCH; urlsToFetchPipe = new Each(urlsToFetchPipe, new CreateUrlDatumFromCrawlDbDatum(maxToFetch)); // Create the sub-assembly that runs the fetch job int maxThreads = isLocal ? CrawlConfig.DEFAULT_NUM_THREADS_LOCAL : CrawlConfig.DEFAULT_NUM_THREADS_CLUSTER; SimpleHttpFetcher fetcher = new SimpleHttpFetcher(maxThreads, fetcherPolicy, userAgent); fetcher.setMaxRetryCount(CrawlConfig.MAX_RETRIES); fetcher.setSocketTimeout(CrawlConfig.SOCKET_TIMEOUT); fetcher.setConnectionTimeout(CrawlConfig.CONNECTION_TIMEOUT); // The scorer is used by the FetchPipe to assign a score to every URL that passes the // robots.txt processing. The score is used to sort URLs such that higher scoring URLs // are fetched first. If URLs are skipped for any reason(s) lower scoring URLs are skipped. BaseScoreGenerator scorer = new FixedScoreGenerator(); FetchPipe fetchPipe = new FetchPipe(urlsToFetchPipe, scorer, fetcher, numReducers); Pipe statusPipe = new Pipe("status pipe", fetchPipe.getStatusTailPipe()); Pipe contentPipe = new Pipe("content pipe", fetchPipe.getContentTailPipe()); contentPipe = TupleLogger.makePipe(contentPipe, true); // Create a parser that returns back the raw HTML (cleaned up by Tika) as the parsed content. SimpleParser parser = new SimpleParser(new ParserPolicy(), true); ParsePipe parsePipe = new ParsePipe(fetchPipe.getContentTailPipe(), parser); Pipe analyzerPipe = new Pipe("analyzer pipe"); analyzerPipe = new Each(parsePipe.getTailPipe(), new AnalyzeHtml()); //add a regex url filter to filter outlinks Pipe outlinksPipe = new Pipe("outlinks pipe", analyzerPipe); outlinksPipe = new Each(outlinksPipe, new CreateLinkDatumFromOutlinksFunction()); if (crawlUrlFilter != null) { outlinksPipe = new Each(outlinksPipe, new UrlFilter(crawlUrlFilter)); } Pipe resultsPipe = new Pipe("results pipe", analyzerPipe); resultsPipe = new Each(resultsPipe, new CreateResultsFunction()); // Group the finished datums, the skipped datums, status, outlinks Pipe updatePipe = new CoGroup("update pipe", Pipe.pipes(finishedDatumsFromDb, statusPipe, analyzerPipe, outlinksPipe), Fields.fields(new Fields(CrawlDbDatum.URL_FIELD), new Fields(StatusDatum.URL_FN), new Fields(AnalyzedDatum.URL_FIELD), new Fields(LinkDatum.URL_FN)), null, new OuterJoin()); updatePipe = new Every(updatePipe, new UpdateCrawlDbBuffer(), Fields.RESULTS); // output : loop dir specific crawldb Path outCrawlDbPath = new Path(curLoopDirPath, CrawlConfig.CRAWLDB_SUBDIR_NAME); Tap crawlDbSink = new Hfs(new SequenceFile(CrawlDbDatum.FIELDS), outCrawlDbPath.toString()); // Status, Path statusDirPath = new Path(curLoopDirPath, CrawlConfig.STATUS_SUBDIR_NAME); Tap statusSink = new Hfs(new TextLine(), statusDirPath.toString()); // Content Path contentDirPath = new Path(curLoopDirPath, CrawlConfig.CONTENT_SUBDIR_NAME); Tap contentSink = new Hfs(new SequenceFile(FetchedDatum.FIELDS), contentDirPath.toString()); // PageResults Path resultsDirPath = new Path(curLoopDirPath, CrawlConfig.RESULTS_SUBDIR_NAME); Tap resultsSink = new Hfs(new TextLine(), resultsDirPath.toString()); // Create the output map that connects each tail pipe to the appropriate sink. Map<String, Tap> sinkMap = new HashMap<String, Tap>(); sinkMap.put(updatePipe.getName(), crawlDbSink); sinkMap.put(statusPipe.getName(), statusSink); sinkMap.put(contentPipe.getName(), contentSink); sinkMap.put(resultsPipe.getName(), resultsSink); FlowConnector flowConnector = new FlowConnector(props); Flow flow = flowConnector.connect(inputSource, sinkMap, updatePipe, statusPipe, contentPipe, resultsPipe); return flow; }
From source file:com.github.gaoyangthu.demo.mapred.dancing.DistributedPentomino.java
License:Apache License
public int run(String[] args) throws Exception { JobConf conf; int depth = 5; int width = 9; int height = 10; Class<? extends Pentomino> pentClass; if (args.length == 0) { System.out.println("Usage: pentomino <output> [-depth #] [-height #] [-width #]"); ToolRunner.printGenericCommandUsage(System.out); return -1; }//from w w w .ja v a 2 s .co m conf = new JobConf(getConf()); // Pick up the parameters, should the user set these width = conf.getInt("pent.width", width); height = conf.getInt("pent.height", height); depth = conf.getInt("pent.depth", depth); pentClass = conf.getClass("pent.class", OneSidedPentomino.class, Pentomino.class); for (int i = 0; i < args.length; i++) { if (args[i].equalsIgnoreCase("-depth")) { depth = Integer.parseInt(args[++i].trim()); } else if (args[i].equalsIgnoreCase("-height")) { height = Integer.parseInt(args[++i].trim()); } else if (args[i].equalsIgnoreCase("-width")) { width = Integer.parseInt(args[++i].trim()); } } // Set parameters for MR tasks to pick up either which way the user sets // them or not conf.setInt("pent.width", width); conf.setInt("pent.height", height); conf.setInt("pent.depth", depth); Path output = new Path(args[0]); Path input = new Path(output + "_input"); FileSystem fileSys = FileSystem.get(conf); try { FileInputFormat.setInputPaths(conf, input); FileOutputFormat.setOutputPath(conf, output); conf.setJarByClass(PentMap.class); conf.setJobName("dancingElephant"); Pentomino pent = ReflectionUtils.newInstance(pentClass, conf); pent.initialize(width, height); createInputDirectory(fileSys, input, pent, depth); // the keys are the prefix strings conf.setOutputKeyClass(Text.class); // the values are puzzle solutions conf.setOutputValueClass(Text.class); conf.setMapperClass(PentMap.class); conf.setReducerClass(IdentityReducer.class); conf.setNumMapTasks(2000); conf.setNumReduceTasks(1); JobClient.runJob(conf); } finally { fileSys.delete(input, true); } return 0; }