List of usage examples for org.apache.hadoop.mapreduce.lib.join CompositeInputFormat JOIN_EXPR
String JOIN_EXPR
To view the source code for org.apache.hadoop.mapreduce.lib.join CompositeInputFormat JOIN_EXPR.
Click Source Link
From source file:com.fanlehai.hadoop.join.CompositeJoin.java
License:Apache License
/** * The main driver for sort program. Invoke this method to submit the * map/reduce job.//from w w w.java 2 s . c o m * * @throws IOException * When there is communication problems with the job tracker. */ @SuppressWarnings("rawtypes") public int run(String[] args) throws Exception { Configuration conf = getConf(); JobClient client = new JobClient(conf); ClusterStatus cluster = client.getClusterStatus(); int num_reduces = (int) (cluster.getMaxReduceTasks() * 0.9); String join_reduces = conf.get(REDUCES_PER_HOST); if (join_reduces != null) { num_reduces = cluster.getTaskTrackers() * Integer.parseInt(join_reduces); } Job job = Job.getInstance(conf); job.setJobName("join"); job.setJarByClass(CompositeJoin.class); job.setMapperClass(Mapper.class); job.setReducerClass(Reducer.class); Class<? extends InputFormat> inputFormatClass = KeyValueTextInputFormat.class;// SequenceFileInputFormat.class; Class<? extends OutputFormat> outputFormatClass = SequenceFileOutputFormat.class; Class<? extends WritableComparable> outputKeyClass = Text.class;// BytesWritable.class; Class<? extends Writable> outputValueClass = Text.class;//TupleWritable.class; String op = "inner"; List<String> otherArgs = new ArrayList<String>(); for (int i = 0; i < args.length; ++i) { try { if ("-r".equals(args[i])) { num_reduces = Integer.parseInt(args[++i]); } else if ("-inFormat".equals(args[i])) { inputFormatClass = Class.forName(args[++i]).asSubclass(InputFormat.class); } else if ("-outFormat".equals(args[i])) { outputFormatClass = Class.forName(args[++i]).asSubclass(OutputFormat.class); } else if ("-outKey".equals(args[i])) { outputKeyClass = Class.forName(args[++i]).asSubclass(WritableComparable.class); } else if ("-outValue".equals(args[i])) { outputValueClass = Class.forName(args[++i]).asSubclass(Writable.class); } else if ("-joinOp".equals(args[i])) { op = args[++i]; } else { otherArgs.add(args[i]); } } catch (NumberFormatException except) { System.out.println("ERROR: Integer expected instead of " + args[i]); return printUsage(); } catch (ArrayIndexOutOfBoundsException except) { System.out.println("ERROR: Required parameter missing from " + args[i - 1]); return printUsage(); // exits } } // Set user-supplied (possibly default) job configs job.setNumReduceTasks(num_reduces); if (otherArgs.size() < 2) { System.out.println("ERROR: Wrong number of parameters: "); return printUsage(); } String strOut = otherArgs.remove(otherArgs.size() - 1); FileSystem.get(new Configuration()).delete(new Path(strOut), true); FileOutputFormat.setOutputPath(job, new Path(strOut)); List<Path> plist = new ArrayList<Path>(otherArgs.size()); for (String s : otherArgs) { plist.add(new Path(s)); } job.setInputFormatClass(CompositeInputFormat.class); job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose(op, inputFormatClass, plist.toArray(new Path[0]))); job.setOutputFormatClass(outputFormatClass); job.setMapperClass(MapComposite.class); job.setOutputKeyClass(outputKeyClass); job.setOutputValueClass(outputValueClass); Date startTime = new Date(); System.out.println("Job started: " + startTime); int ret = job.waitForCompletion(true) ? 0 : 1; Date end_time = new Date(); System.out.println("Job ended: " + end_time); System.out.println("The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds."); return ret; }
From source file:com.phantom.hadoop.examples.Join.java
License:Apache License
/** * The main driver for sort program. Invoke this method to submit the * map/reduce job./*from w ww . j a va 2 s . c o m*/ * * @throws IOException * When there is communication problems with the job tracker. */ @SuppressWarnings("unchecked") public int run(String[] args) throws Exception { Configuration conf = getConf(); JobClient client = new JobClient(conf); ClusterStatus cluster = client.getClusterStatus(); int num_reduces = (int) (cluster.getMaxReduceTasks() * 0.9); String join_reduces = conf.get(REDUCES_PER_HOST); if (join_reduces != null) { num_reduces = cluster.getTaskTrackers() * Integer.parseInt(join_reduces); } Job job = new Job(conf); job.setJobName("join"); job.setJarByClass(Sort.class); job.setMapperClass(Mapper.class); job.setReducerClass(Reducer.class); Class<? extends InputFormat> inputFormatClass = SequenceFileInputFormat.class; Class<? extends OutputFormat> outputFormatClass = SequenceFileOutputFormat.class; Class<? extends WritableComparable> outputKeyClass = BytesWritable.class; Class<? extends Writable> outputValueClass = TupleWritable.class; String op = "inner"; List<String> otherArgs = new ArrayList<String>(); for (int i = 0; i < args.length; ++i) { try { if ("-r".equals(args[i])) { num_reduces = Integer.parseInt(args[++i]); } else if ("-inFormat".equals(args[i])) { inputFormatClass = Class.forName(args[++i]).asSubclass(InputFormat.class); } else if ("-outFormat".equals(args[i])) { outputFormatClass = Class.forName(args[++i]).asSubclass(OutputFormat.class); } else if ("-outKey".equals(args[i])) { outputKeyClass = Class.forName(args[++i]).asSubclass(WritableComparable.class); } else if ("-outValue".equals(args[i])) { outputValueClass = Class.forName(args[++i]).asSubclass(Writable.class); } else if ("-joinOp".equals(args[i])) { op = args[++i]; } else { otherArgs.add(args[i]); } } catch (NumberFormatException except) { System.out.println("ERROR: Integer expected instead of " + args[i]); return printUsage(); } catch (ArrayIndexOutOfBoundsException except) { System.out.println("ERROR: Required parameter missing from " + args[i - 1]); return printUsage(); // exits } } // Set user-supplied (possibly default) job configs job.setNumReduceTasks(num_reduces); if (otherArgs.size() < 2) { System.out.println("ERROR: Wrong number of parameters: "); return printUsage(); } FileOutputFormat.setOutputPath(job, new Path(otherArgs.remove(otherArgs.size() - 1))); List<Path> plist = new ArrayList<Path>(otherArgs.size()); for (String s : otherArgs) { plist.add(new Path(s)); } job.setInputFormatClass(CompositeInputFormat.class); job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose(op, inputFormatClass, plist.toArray(new Path[0]))); job.setOutputFormatClass(outputFormatClass); job.setOutputKeyClass(outputKeyClass); job.setOutputValueClass(outputValueClass); Date startTime = new Date(); System.out.println("Job started: " + startTime); int ret = job.waitForCompletion(true) ? 0 : 1; Date end_time = new Date(); System.out.println("Job ended: " + end_time); System.out.println("The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds."); return ret; }
From source file:com.twitter.algebra.matrix.multiply.AtBOuterStaticMapsideJoinJob.java
License:Apache License
public void run(Configuration conf, Path atPath, Path bPath, Path outPath, int outCardinality) throws IOException, InterruptedException, ClassNotFoundException { conf.setInt(OUT_CARD, outCardinality); @SuppressWarnings("deprecation") Job job = new Job(conf); job.setJobName(AtBOuterStaticMapsideJoinJob.class.getSimpleName()); job.setJarByClass(AtBOuterStaticMapsideJoinJob.class); FileSystem fs = FileSystem.get(atPath.toUri(), conf); atPath = fs.makeQualified(atPath);/* w ww . j a va 2 s. co m*/ bPath = fs.makeQualified(bPath); job.setInputFormatClass(CompositeInputFormat.class); //mapside join expression job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose("inner", SequenceFileInputFormat.class, atPath, bPath)); job.setOutputFormatClass(MatrixOutputFormat.class); outPath = fs.makeQualified(outPath); FileOutputFormat.setOutputPath(job, outPath); job.setMapperClass(MyMapper.class); job.setMapOutputKeyClass(IntWritable.class); job.setMapOutputValueClass(VectorWritable.class); job.setCombinerClass(MyReducer.class); int numReducers = conf.getInt("algebra.reduceslots.multiply", 10); job.setNumReduceTasks(numReducers); job.setReducerClass(MyReducer.class); job.setOutputKeyClass(IntWritable.class); job.setOutputValueClass(VectorWritable.class); job.submit(); boolean res = job.waitForCompletion(true); if (!res) throw new IOException("Job failed"); }
From source file:org.apache.hadoop.examples.Join.java
License:Apache License
/** * The main driver for sort program./*w w w . j a v a2s .c om*/ * Invoke this method to submit the map/reduce job. * @throws IOException When there is communication problems with the * job tracker. */ @SuppressWarnings("unchecked") public int run(String[] args) throws Exception { Configuration conf = getConf(); JobClient client = new JobClient(conf); ClusterStatus cluster = client.getClusterStatus(); int num_reduces = (int) (cluster.getMaxReduceTasks() * 0.9); String join_reduces = conf.get(REDUCES_PER_HOST); if (join_reduces != null) { num_reduces = cluster.getTaskTrackers() * Integer.parseInt(join_reduces); } Job job = Job.getInstance(conf); job.setJobName("join"); job.setJarByClass(Sort.class); job.setMapperClass(Mapper.class); job.setReducerClass(Reducer.class); Class<? extends InputFormat> inputFormatClass = SequenceFileInputFormat.class; Class<? extends OutputFormat> outputFormatClass = SequenceFileOutputFormat.class; Class<? extends WritableComparable> outputKeyClass = BytesWritable.class; Class<? extends Writable> outputValueClass = TupleWritable.class; String op = "inner"; List<String> otherArgs = new ArrayList<String>(); for (int i = 0; i < args.length; ++i) { try { if ("-r".equals(args[i])) { num_reduces = Integer.parseInt(args[++i]); } else if ("-inFormat".equals(args[i])) { inputFormatClass = Class.forName(args[++i]).asSubclass(InputFormat.class); } else if ("-outFormat".equals(args[i])) { outputFormatClass = Class.forName(args[++i]).asSubclass(OutputFormat.class); } else if ("-outKey".equals(args[i])) { outputKeyClass = Class.forName(args[++i]).asSubclass(WritableComparable.class); } else if ("-outValue".equals(args[i])) { outputValueClass = Class.forName(args[++i]).asSubclass(Writable.class); } else if ("-joinOp".equals(args[i])) { op = args[++i]; } else { otherArgs.add(args[i]); } } catch (NumberFormatException except) { System.out.println("ERROR: Integer expected instead of " + args[i]); return printUsage(); } catch (ArrayIndexOutOfBoundsException except) { System.out.println("ERROR: Required parameter missing from " + args[i - 1]); return printUsage(); // exits } } // Set user-supplied (possibly default) job configs job.setNumReduceTasks(num_reduces); if (otherArgs.size() < 2) { System.out.println("ERROR: Wrong number of parameters: "); return printUsage(); } FileOutputFormat.setOutputPath(job, new Path(otherArgs.remove(otherArgs.size() - 1))); List<Path> plist = new ArrayList<Path>(otherArgs.size()); for (String s : otherArgs) { plist.add(new Path(s)); } job.setInputFormatClass(CompositeInputFormat.class); job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose(op, inputFormatClass, plist.toArray(new Path[0]))); job.setOutputFormatClass(outputFormatClass); job.setOutputKeyClass(outputKeyClass); job.setOutputValueClass(outputValueClass); Date startTime = new Date(); System.out.println("Job started: " + startTime); int ret = job.waitForCompletion(true) ? 0 : 1; Date end_time = new Date(); System.out.println("Job ended: " + end_time); System.out.println("The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds."); return ret; }
From source file:org.apache.Join.java
License:Apache License
/** * The main driver for sort program./*from www . j av a2s.c om*/ * Invoke this method to submit the map/reduce job. * @throws IOException When there is communication problems with the * job tracker. */ @SuppressWarnings("unchecked") public int run(String[] args) throws Exception { Configuration conf = getConf(); JobClient client = new JobClient(conf); ClusterStatus cluster = client.getClusterStatus(); int num_reduces = (int) (cluster.getMaxReduceTasks() * 0.9); String join_reduces = conf.get(REDUCES_PER_HOST); if (join_reduces != null) { num_reduces = cluster.getTaskTrackers() * Integer.parseInt(join_reduces); } Job job = new Job(conf); job.setJobName("join"); job.setJarByClass(Sort.class); job.setMapperClass(Mapper.class); job.setReducerClass(Reducer.class); Class<? extends InputFormat> inputFormatClass = KeyValueTextInputFormat.class; Class<? extends OutputFormat> outputFormatClass = SequenceFileOutputFormat.class; Class<? extends WritableComparable> outputKeyClass = BytesWritable.class; Class<? extends Writable> outputValueClass = TupleWritable.class; String op = "inner"; List<String> otherArgs = new ArrayList<String>(); for (int i = 0; i < args.length; ++i) { try { if ("-r".equals(args[i])) { num_reduces = Integer.parseInt(args[++i]); } else if ("-inFormat".equals(args[i])) { inputFormatClass = Class.forName(args[++i]).asSubclass(InputFormat.class); } else if ("-outFormat".equals(args[i])) { outputFormatClass = Class.forName(args[++i]).asSubclass(OutputFormat.class); } else if ("-outKey".equals(args[i])) { outputKeyClass = Class.forName(args[++i]).asSubclass(WritableComparable.class); } else if ("-outValue".equals(args[i])) { outputValueClass = Class.forName(args[++i]).asSubclass(Writable.class); } else if ("-joinOp".equals(args[i])) { op = args[++i]; } else { otherArgs.add(args[i]); } } catch (NumberFormatException except) { System.out.println("ERROR: Integer expected instead of " + args[i]); return printUsage(); } catch (ArrayIndexOutOfBoundsException except) { System.out.println("ERROR: Required parameter missing from " + args[i - 1]); return printUsage(); // exits } } // Set user-supplied (possibly default) job configs job.setNumReduceTasks(num_reduces); if (otherArgs.size() < 2) { System.out.println("ERROR: Wrong number of parameters: "); return printUsage(); } FileOutputFormat.setOutputPath(job, new Path(otherArgs.remove(otherArgs.size() - 1))); List<Path> plist = new ArrayList<Path>(otherArgs.size()); for (String s : otherArgs) { plist.add(new Path(s)); } job.setInputFormatClass(CompositeInputFormat.class); job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose(op, inputFormatClass, plist.toArray(new Path[0]))); job.setOutputFormatClass(outputFormatClass); job.setOutputKeyClass(outputKeyClass); job.setOutputValueClass(outputValueClass); Date startTime = new Date(); System.out.println("Job started: " + startTime); int ret = job.waitForCompletion(true) ? 0 : 1; Date end_time = new Date(); System.out.println("Job ended: " + end_time); System.out.println("The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds."); return ret; }
From source file:org.apache.tez.mapreduce.examples.Join.java
License:Apache License
/** * The main driver for sort program.//from w w w.j av a 2 s .c om * Invoke this method to submit the map/reduce job. * @throws Exception When there is communication problems with the * job tracker. */ public int run(String[] args) throws Exception { Configuration conf = getConf(); JobClient client = new JobClient(conf); ClusterStatus cluster = client.getClusterStatus(); int num_reduces = (int) (cluster.getMaxReduceTasks() * 0.9); String join_reduces = conf.get(REDUCES_PER_HOST); if (join_reduces != null) { num_reduces = cluster.getTaskTrackers() * Integer.parseInt(join_reduces); } Job job = new Job(conf); job.setJobName("join"); job.setJarByClass(Sort.class); job.setMapperClass(Mapper.class); job.setReducerClass(Reducer.class); Class<? extends InputFormat> inputFormatClass = SequenceFileInputFormat.class; Class<? extends OutputFormat> outputFormatClass = SequenceFileOutputFormat.class; Class<? extends WritableComparable> outputKeyClass = BytesWritable.class; Class<? extends Writable> outputValueClass = TupleWritable.class; String op = "inner"; List<String> otherArgs = new ArrayList<String>(); for (int i = 0; i < args.length; ++i) { try { if ("-r".equals(args[i])) { num_reduces = Integer.parseInt(args[++i]); } else if ("-inFormat".equals(args[i])) { inputFormatClass = Class.forName(args[++i]).asSubclass(InputFormat.class); } else if ("-outFormat".equals(args[i])) { outputFormatClass = Class.forName(args[++i]).asSubclass(OutputFormat.class); } else if ("-outKey".equals(args[i])) { outputKeyClass = Class.forName(args[++i]).asSubclass(WritableComparable.class); } else if ("-outValue".equals(args[i])) { outputValueClass = Class.forName(args[++i]).asSubclass(Writable.class); } else if ("-joinOp".equals(args[i])) { op = args[++i]; } else { otherArgs.add(args[i]); } } catch (NumberFormatException except) { System.out.println("ERROR: Integer expected instead of " + args[i]); return printUsage(); } catch (ArrayIndexOutOfBoundsException except) { System.out.println("ERROR: Required parameter missing from " + args[i - 1]); return printUsage(); // exits } } // Set user-supplied (possibly default) job configs job.setNumReduceTasks(num_reduces); if (otherArgs.size() < 2) { System.out.println("ERROR: Wrong number of parameters: "); return printUsage(); } FileOutputFormat.setOutputPath(job, new Path(otherArgs.remove(otherArgs.size() - 1))); List<Path> plist = new ArrayList<Path>(otherArgs.size()); for (String s : otherArgs) { plist.add(new Path(s)); } job.setInputFormatClass(CompositeInputFormat.class); job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose(op, inputFormatClass, plist.toArray(new Path[0]))); job.setOutputFormatClass(outputFormatClass); job.setOutputKeyClass(outputKeyClass); job.setOutputValueClass(outputValueClass); Date startTime = new Date(); System.out.println("Job started: " + startTime); int ret = job.waitForCompletion(true) ? 0 : 1; Date end_time = new Date(); System.out.println("Job ended: " + end_time); System.out.println("The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds."); return ret; }
From source file:seoeun.hadoop.dbquery.join.Join2.java
License:Apache License
/** * The main driver for sort program. Invoke this method to submit the map/reduce job. * /*from w w w . j ava 2 s . com*/ * @throws IOException When there is communication problems with the job tracker. */ @SuppressWarnings("unchecked") public int run(String[] args) throws Exception { System.out.println(" ------- run"); Configuration conf = getConfiguration(); JobClient client = new JobClient(conf); ClusterStatus cluster = client.getClusterStatus(); int num_reduces = (int) (cluster.getMaxReduceTasks() * 0.9); String join_reduces = conf.get(REDUCES_PER_HOST); if (join_reduces != null) { num_reduces = cluster.getTaskTrackers() * Integer.parseInt(join_reduces); } Job job = new Job(conf); job.setJobName("join"); job.setJarByClass(Sort.class); job.setMapperClass(Mapper.class); job.setReducerClass(Reducer.class); Class<? extends InputFormat> inputFormatClass = KeyValueTextInputFormat.class; Class<? extends OutputFormat> outputFormatClass = TextOutputFormat.class; Class<? extends WritableComparable> outputKeyClass = Text.class; Class<? extends Writable> outputValueClass = TupleWritable.class; String op = "inner"; List<String> otherArgs = new ArrayList<String>(); for (int i = 0; i < args.length; ++i) { try { if ("-r".equals(args[i])) { num_reduces = Integer.parseInt(args[++i]); } else if ("-inFormat".equals(args[i])) { inputFormatClass = Class.forName(args[++i]).asSubclass(InputFormat.class); } else if ("-outFormat".equals(args[i])) { outputFormatClass = Class.forName(args[++i]).asSubclass(OutputFormat.class); } else if ("-outKey".equals(args[i])) { outputKeyClass = Class.forName(args[++i]).asSubclass(WritableComparable.class); } else if ("-outValue".equals(args[i])) { outputValueClass = Class.forName(args[++i]).asSubclass(Writable.class); } else if ("-joinOp".equals(args[i])) { op = args[++i]; } else { otherArgs.add(args[i]); } } catch (NumberFormatException except) { System.out.println("ERROR: Integer expected instead of " + args[i]); return printUsage(); } catch (ArrayIndexOutOfBoundsException except) { System.out.println("ERROR: Required parameter missing from " + args[i - 1]); return printUsage(); // exits } } // Set user-supplied (possibly default) job configs job.setNumReduceTasks(num_reduces); if (otherArgs.size() < 2) { System.out.println("ERROR: Wrong number of parameters: "); return printUsage(); } FileOutputFormat.setOutputPath(job, new Path(otherArgs.remove(otherArgs.size() - 1))); List<Path> plist = new ArrayList<Path>(otherArgs.size()); for (String s : otherArgs) { plist.add(new Path(s)); } job.setInputFormatClass(CompositeInputFormat.class); job.getConfiguration().set(CompositeInputFormat.JOIN_EXPR, CompositeInputFormat.compose(op, inputFormatClass, plist.toArray(new Path[0]))); job.setOutputFormatClass(outputFormatClass); job.setOutputKeyClass(outputKeyClass); job.setOutputValueClass(outputValueClass); Date startTime = new Date(); System.out.println("Job started: " + startTime); int ret = job.waitForCompletion(true) ? 0 : 1; Date end_time = new Date(); System.out.println("Job ended: " + end_time); System.out.println("The job took " + (end_time.getTime() - startTime.getTime()) / 1000 + " seconds."); return ret; }