List of usage examples for org.apache.hadoop.mapreduce.lib.partition TotalOrderPartitioner getPartitionFile
public static String getPartitionFile(Configuration conf)
From source file:ComRoughSetApproInputSampler.java
License:Apache License
/** * Write a partition file for the given job, using the Sampler provided. * Queries the sampler for a sample keyset, sorts by the output key * comparator, selects the keys for each rank, and writes to the destination * returned from {@link TotalOrderPartitioner#getPartitionFile}. *//*from ww w. j a v a2 s . co m*/ @SuppressWarnings("unchecked") // getInputFormat, getOutputKeyComparator public static <K, V> void writePartitionFile(Job job, Sampler<K, V> sampler) throws IOException, ClassNotFoundException, InterruptedException { Configuration conf = job.getConfiguration(); final InputFormat inf = ReflectionUtils.newInstance(job.getInputFormatClass(), conf); int numPartitions = job.getNumReduceTasks(); K[] samples = (K[]) sampler.getSample(inf, job); LOG.info("Using " + samples.length + " samples"); RawComparator<K> comparator = (RawComparator<K>) job.getSortComparator(); Arrays.sort(samples, comparator); Path dst = new Path(TotalOrderPartitioner.getPartitionFile(conf)); FileSystem fs = dst.getFileSystem(conf); if (fs.exists(dst)) { fs.delete(dst, false); } SequenceFile.Writer writer = SequenceFile.createWriter(fs, conf, dst, job.getMapOutputKeyClass(), NullWritable.class); NullWritable nullValue = NullWritable.get(); float stepSize = samples.length / (float) numPartitions; int last = -1; for (int i = 1; i < numPartitions; ++i) { int k = Math.round(stepSize * i); while (last >= k && comparator.compare(samples[last], samples[k]) == 0) { ++k; } writer.append(samples[k], nullValue); last = k; } writer.close(); }
From source file:com.savy3.nonequijoin.MapOutputSampler.java
License:Apache License
/** * Write a partition file for the given job, using the Sampler provided. * Queries the sampler for a sample keyset, sorts by the output key * comparator, selects the keys for each rank, and writes to the destination * returned from {@link TotalOrderPartitioner#getPartitionFile}. *///from www . j a v a2 s .co m @SuppressWarnings("unchecked") // getInputFormat, getOutputKeyComparator public static <K, V> void writePartitionFile(Job job, Sampler<K, V> sampler) throws IOException, ClassNotFoundException, InterruptedException { Configuration conf = job.getConfiguration(); final InputFormat inf = ReflectionUtils.newInstance(job.getInputFormatClass(), conf); int numPartitions = job.getNumReduceTasks(); HashMap<K, V> samples = (HashMap<K, V>) sampler.getSample(inf, job); LOG.info("Using " + samples.size() + " samples"); // write the input samples in to file <partitionfile>/mapIn Path dstOut = new Path(TotalOrderPartitioner.getPartitionFile(conf)); Path dst = new Path(dstOut, "mapIn"); FileSystem fs = dst.getFileSystem(conf); SequenceFile.Writer sampleWriter = null; for (Map.Entry<K, V> sample : samples.entrySet()) { sampleWriter = SequenceFile.createWriter(fs, conf, dst, sample.getKey().getClass(), sample.getValue().getClass()); break; } for (Map.Entry<K, V> sample : samples.entrySet()) { sampleWriter.append(sample.getKey(), sample.getValue()); } sampleWriter.close(); LOG.info("Sample Input File location " + dst.toString()); // run map reduce on the samples input runMap(job, dst); }
From source file:crunch.MaxTemperature.java
License:Apache License
@Override
public int run(String[] args) throws Exception {
Job job = JobBuilder.parseInputAndOutput(this, getConf(), args);
if (job == null) {
return -1;
}/*from www . j a va 2 s . com*/
job.setInputFormatClass(SequenceFileInputFormat.class);
job.setOutputKeyClass(IntWritable.class);
job.setOutputFormatClass(SequenceFileOutputFormat.class);
SequenceFileOutputFormat.setCompressOutput(job, true);
SequenceFileOutputFormat.setOutputCompressorClass(job, GzipCodec.class);
SequenceFileOutputFormat.setOutputCompressionType(job, CompressionType.BLOCK);
job.setPartitionerClass(TotalOrderPartitioner.class);
InputSampler.Sampler<IntWritable, Text> sampler = new InputSampler.RandomSampler<IntWritable, Text>(0.1,
10000, 10);
InputSampler.writePartitionFile(job, sampler);
// Add to DistributedCache
Configuration conf = job.getConfiguration();
String partitionFile = TotalOrderPartitioner.getPartitionFile(conf);
URI partitionUri = new URI(partitionFile + "#" + TotalOrderPartitioner.DEFAULT_PATH);
DistributedCache.addCacheFile(partitionUri, conf);
DistributedCache.createSymlink(conf);
return job.waitForCompletion(true) ? 0 : 1;
}
From source file:org.rdfhdt.mrbuilder.HDTBuilderDriver.java
License:Open Source License
protected boolean runDictionaryJob() throws ClassNotFoundException, IOException, InterruptedException, URISyntaxException { boolean jobOK; Job job = null;/*from w ww . jav a 2 s .c o m*/ BufferedWriter bufferedWriter; // if output path exists... if (this.dictionaryFS.exists(this.conf.getDictionaryOutputPath())) { if (this.conf.getDeleteDictionaryOutputPath()) { // ... and option provided, delete recursively this.dictionaryFS.delete(this.conf.getDictionaryOutputPath(), true); } else { // ... and option not provided, fail System.out.println("Dictionary output path does exist: " + this.conf.getDictionaryOutputPath()); System.out.println("Select other path or use option -dd to overwrite"); System.exit(-1); } } // Sample the SequenceInputFormat to do TotalSort and create final output job = new Job(this.conf.getConfigurationObject(), this.conf.getDictionaryJobName() + " phase 2"); job.setJarByClass(HDTBuilderDriver.class); System.out.println("samples = " + this.conf.getDictionarySamplesPath()); System.out.println("output = " + this.conf.getDictionaryOutputPath()); FileInputFormat.addInputPath(job, this.conf.getDictionarySamplesPath()); FileOutputFormat.setOutputPath(job, this.conf.getDictionaryOutputPath()); job.setInputFormatClass(SequenceFileInputFormat.class); LazyOutputFormat.setOutputFormatClass(job, SequenceFileOutputFormat.class); // Identity Mapper // job.setMapperClass(Mapper.class); job.setCombinerClass(DictionaryCombiner.class); job.setPartitionerClass(TotalOrderPartitioner.class); job.setReducerClass(DictionaryReducer.class); job.setNumReduceTasks(this.conf.getDictionaryReducers()); job.setMapOutputKeyClass(Text.class); job.setMapOutputValueClass(Text.class); job.setOutputKeyClass(Text.class); job.setOutputValueClass(NullWritable.class); System.out.println("Sampling started"); InputSampler.writePartitionFile(job, new InputSampler.IntervalSampler<Text, Text>(this.conf.getSampleProbability())); String partitionFile = TotalOrderPartitioner.getPartitionFile(job.getConfiguration()); URI partitionUri = new URI(partitionFile + "#" + TotalOrderPartitioner.DEFAULT_PATH); DistributedCache.addCacheFile(partitionUri, job.getConfiguration()); DistributedCache.createSymlink(job.getConfiguration()); System.out.println("Sampling finished"); MultipleOutputs.addNamedOutput(job, HDTBuilderConfiguration.SHARED, SequenceFileOutputFormat.class, Text.class, NullWritable.class); MultipleOutputs.addNamedOutput(job, HDTBuilderConfiguration.SUBJECTS, SequenceFileOutputFormat.class, Text.class, NullWritable.class); MultipleOutputs.addNamedOutput(job, HDTBuilderConfiguration.PREDICATES, SequenceFileOutputFormat.class, Text.class, NullWritable.class); MultipleOutputs.addNamedOutput(job, HDTBuilderConfiguration.OBJECTS, SequenceFileOutputFormat.class, Text.class, NullWritable.class); SequenceFileOutputFormat.setCompressOutput(job, true); SequenceFileOutputFormat.setOutputCompressorClass(job, com.hadoop.compression.lzo.LzoCodec.class); SequenceFileOutputFormat.setOutputCompressionType(job, SequenceFile.CompressionType.BLOCK); jobOK = job.waitForCompletion(true); this.numShared = job.getCounters().findCounter(Counters.Shared).getValue(); this.numSubjects = job.getCounters().findCounter(Counters.Subjects).getValue(); this.numPredicates = job.getCounters().findCounter(Counters.Predicates).getValue(); this.numObjects = job.getCounters().findCounter(Counters.Objects).getValue(); bufferedWriter = new BufferedWriter( new OutputStreamWriter(this.dictionaryFS.create(this.conf.getDictionaryCountersFile()))); bufferedWriter.write(HDTBuilderConfiguration.SHARED + "=" + this.numShared + "\n"); bufferedWriter.write(HDTBuilderConfiguration.SUBJECTS + "=" + this.numSubjects + "\n"); bufferedWriter.write(HDTBuilderConfiguration.PREDICATES + "=" + this.numPredicates + "\n"); bufferedWriter.write(HDTBuilderConfiguration.OBJECTS + "=" + this.numObjects + "\n"); bufferedWriter.close(); return jobOK; }
From source file:org.rdfhdt.mrbuilder.HDTBuilderDriver.java
License:Open Source License
protected boolean runTriplesJob() throws IOException, ClassNotFoundException, InterruptedException, URISyntaxException { Job job = null;/* w ww. j a v a2s. c o m*/ boolean jobOK; // if triples output path exists... if (this.triplesFS.exists(this.conf.getTriplesOutputPath())) { if (this.conf.getDeleteTriplesOutputPath()) { // ... and option provided, delete recursively this.triplesFS.delete(this.conf.getTriplesOutputPath(), true); } else { // ... and option not provided, fail System.out.println("Triples output path does exist: " + this.conf.getTriplesOutputPath()); System.out.println("Select other path or use option -dt to overwrite"); System.exit(-1); } } job = new Job(this.conf.getConfigurationObject(), this.conf.getTriplesJobName() + " phase 2"); job.setJarByClass(HDTBuilderDriver.class); FileInputFormat.addInputPath(job, this.conf.getTriplesSamplesPath()); FileOutputFormat.setOutputPath(job, this.conf.getTriplesOutputPath()); job.setInputFormatClass(SequenceFileInputFormat.class); LazyOutputFormat.setOutputFormatClass(job, SequenceFileOutputFormat.class); job.setSortComparatorClass(TripleSPOComparator.class); job.setGroupingComparatorClass(TripleSPOComparator.class); job.setPartitionerClass(TotalOrderPartitioner.class); job.setOutputKeyClass(TripleSPOWritable.class); job.setOutputValueClass(NullWritable.class); job.setNumReduceTasks(this.conf.getTriplesReducers()); System.out.println("Sampling started"); InputSampler.writePartitionFile(job, new InputSampler.IntervalSampler<Text, Text>(this.conf.getSampleProbability())); String partitionFile = TotalOrderPartitioner.getPartitionFile(job.getConfiguration()); URI partitionUri = new URI(partitionFile + "#" + TotalOrderPartitioner.DEFAULT_PATH); DistributedCache.addCacheFile(partitionUri, job.getConfiguration()); DistributedCache.createSymlink(job.getConfiguration()); System.out.println("Sampling finished"); SequenceFileOutputFormat.setCompressOutput(job, true); SequenceFileOutputFormat.setOutputCompressorClass(job, com.hadoop.compression.lzo.LzoCodec.class); SequenceFileOutputFormat.setOutputCompressionType(job, SequenceFile.CompressionType.BLOCK); jobOK = job.waitForCompletion(true); return jobOK; }