List of usage examples for org.apache.hadoop.mapreduce JobID JobID
public JobID()
From source file:org.apache.pig.backend.hadoop.executionengine.fetch.FetchPOStoreImpl.java
License:Apache License
@Override public StoreFuncInterface createStoreFunc(POStore store) throws IOException { Configuration conf = ConfigurationUtil.toConfiguration(pc.getProperties()); StoreFuncInterface storeFunc = store.getStoreFunc(); JobContext jc = HadoopShims.createJobContext(conf, new JobID()); OutputFormat<?, ?> outputFormat = storeFunc.getOutputFormat(); PigOutputFormat.setLocation(jc, store); context = HadoopShims.createTaskAttemptContext(conf, HadoopShims.getNewTaskAttemptID()); PigOutputFormat.setLocation(context, store); try {/* w ww .ja v a 2 s.co m*/ outputFormat.checkOutputSpecs(jc); } catch (InterruptedException e) { throw new IOException(e); } try { outputCommitter = outputFormat.getOutputCommitter(context); outputCommitter.setupJob(jc); outputCommitter.setupTask(context); writer = outputFormat.getRecordWriter(context); } catch (InterruptedException e) { throw new IOException(e); } storeFunc.prepareToWrite(writer); return storeFunc; }
From source file:org.apache.pig.impl.io.PigFile.java
License:Apache License
public void store(DataBag data, FuncSpec storeFuncSpec, PigContext pigContext) throws IOException { Configuration conf = ConfigurationUtil.toConfiguration(pigContext.getProperties()); // create a simulated JobContext JobContext jc = HadoopShims.createJobContext(conf, new JobID()); StoreFuncInterface sfunc = (StoreFuncInterface) PigContext.instantiateFuncFromSpec(storeFuncSpec); OutputFormat<?, ?> of = sfunc.getOutputFormat(); POStore store = new POStore(new OperatorKey()); store.setSFile(new FileSpec(file, storeFuncSpec)); PigOutputFormat.setLocation(jc, store); OutputCommitter oc;// w w w . j av a2 s. c om // create a simulated TaskAttemptContext TaskAttemptContext tac = HadoopShims.createTaskAttemptContext(conf, HadoopShims.getNewTaskAttemptID()); PigOutputFormat.setLocation(tac, store); RecordWriter<?, ?> rw; try { of.checkOutputSpecs(jc); oc = of.getOutputCommitter(tac); oc.setupJob(jc); oc.setupTask(tac); rw = of.getRecordWriter(tac); sfunc.prepareToWrite(rw); for (Iterator<Tuple> it = data.iterator(); it.hasNext();) { Tuple row = it.next(); sfunc.putNext(row); } rw.close(tac); } catch (InterruptedException e) { throw new IOException(e); } if (oc.needsTaskCommit(tac)) { oc.commitTask(tac); } HadoopShims.commitOrCleanup(oc, jc); }
From source file:org.apache.pig.impl.io.ReadToEndLoader.java
License:Apache License
@SuppressWarnings("unchecked") private void init() throws IOException { if (conf != null && pigContext != null) { SchemaTupleBackend.initialize(conf, pigContext, true); }// ww w.j a v a 2 s . com // make a copy so that if the underlying InputFormat writes to the // conf, we don't affect the caller's copy conf = new Configuration(conf); // let's initialize the wrappedLoadFunc Job job = new Job(conf); wrappedLoadFunc.setUDFContextSignature(this.udfContextSignature); wrappedLoadFunc.setLocation(inputLocation, job); // The above setLocation call could write to the conf within // the job - get a hold of the modified conf conf = job.getConfiguration(); inputFormat = wrappedLoadFunc.getInputFormat(); try { inpSplits = inputFormat.getSplits(HadoopShims.createJobContext(conf, new JobID())); } catch (InterruptedException e) { throw new IOException(e); } }
From source file:org.goldenorb.OrbPartition.java
License:Apache License
private void dumpData() { Configuration conf = new Configuration(); Job job = null;/* www. j ava 2 s . c o m*/ JobContext jobContext = null; TaskAttemptContext tao = null; RecordWriter rw; VertexWriter vw; FileOutputFormat outputFormat; boolean tryAgain = true; int count = 0; while (tryAgain && count < 15) try { count++; tryAgain = false; if (job == null) { job = new Job(conf); job.setOutputFormatClass(TextOutputFormat.class); FileOutputFormat.setOutputPath(job, new Path(new String(getOrbConf().getNameNode() + getOrbConf().getFileOutputPath()))); } if (jobContext == null) { jobContext = new JobContext(job.getConfiguration(), new JobID()); } System.out.println(jobContext.getConfiguration().get("mapred.output.dir")); tao = new TaskAttemptContext(jobContext.getConfiguration(), new TaskAttemptID(new TaskID(jobContext.getJobID(), true, getPartitionID()), 0)); outputFormat = (FileOutputFormat) tao.getOutputFormatClass().newInstance(); rw = outputFormat.getRecordWriter(tao); vw = (VertexWriter) getOrbConf().getVertexOutputFormatClass().newInstance(); for (Vertex v : vertices.values()) { OrbContext oc = vw.vertexWrite(v); rw.write(oc.getKey(), oc.getValue()); // orbLogger.info("Partition: " + Integer.toString(partitionId) + "writing: " + // oc.getKey().toString() + ", " + oc.getValue().toString()); } rw.close(tao); FileOutputCommitter cm = (FileOutputCommitter) outputFormat.getOutputCommitter(tao); if (cm.needsTaskCommit(tao)) { cm.commitTask(tao); cm.cleanupJob(jobContext); } else { cm.cleanupJob(jobContext); tryAgain = true; } } catch (IOException e) { tryAgain = true; e.printStackTrace(); } catch (InstantiationException e) { tryAgain = true; e.printStackTrace(); } catch (IllegalAccessException e) { tryAgain = true; e.printStackTrace(); } catch (ClassNotFoundException e) { tryAgain = true; e.printStackTrace(); } catch (InterruptedException e) { tryAgain = true; e.printStackTrace(); } if (tryAgain) { synchronized (this) { try { wait(1000); } catch (InterruptedException e) { e.printStackTrace(); } } } }
From source file:org.slc.sli.aggregation.mapreduce.io.MockTaskAttemptContext.java
License:Apache License
@Override public JobID getJobID() { return new JobID(); }