Package org.apache.tez.dag.api

Examples of org.apache.tez.dag.api.InputDescriptor

153
154
155
156
157
158
159
160
161
      this.assigned = assignedSize;
    }
  }

  private InputDescriptor createTestInputDescriptor(Class<? extends LogicalInput> inputClazz) {
    InputDescriptor desc = mock(InputDescriptor.class);
    doReturn(inputClazz.getName()).when(desc).getClassName();
    return desc;
  }
View Full Code Here
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
   
    Path mapInput = new Path(workDir, "map0");
    MapUtils.generateInputSplit(localFs, workDir, mapConf, mapInput);
   
    InputSpec mapInputSpec = new InputSpec("NullSrcVertex",
        new InputDescriptor(MRInputLegacy.class.getName())
            .setUserPayload(MRHelpers.createMRInputPayload(mapConf, null)),
        0);
    OutputSpec mapOutputSpec = new OutputSpec("NullDestVertex", new OutputDescriptor(LocalOnFileSorterOutput.class.getName()), 1);
    // Run a map
    LogicalIOProcessorRuntimeTask mapTask = MapUtils.createLogicalTask(localFs, workDir, mapConf, 0,
        mapInput, new TestUmbilical(), dagName, mapVertexName,
        Collections.singletonList(mapInputSpec),
        Collections.singletonList(mapOutputSpec));

    mapTask.initialize();
    mapTask.run();
    mapTask.close();
   
    LOG.info("Starting reduce...");
   
    Token<JobTokenIdentifier> shuffleToken = new Token<JobTokenIdentifier>();
   
    Configuration reduceStageConf = MultiStageMRConfigUtil.getConfForVertex(conf,
        reduceVertexName);
    JobConf reduceConf = new JobConf(reduceStageConf);
    reduceConf.setOutputFormat(SequenceFileOutputFormat.class);
    reduceConf.set(TezJobConfig.TASK_LOCAL_RESOURCE_DIR, new Path(workDir,
        "localized-resources").toUri().toString());
    FileOutputFormat.setOutputPath(reduceConf, new Path(workDir, "output"));
    ProcessorDescriptor reduceProcessorDesc = new ProcessorDescriptor(
        ReduceProcessor.class.getName()).setUserPayload(TezUtils.createUserPayloadFromConf(reduceConf));
   
    InputSpec reduceInputSpec = new InputSpec(mapVertexName,
        new InputDescriptor(LocalMergedInput.class.getName()), 1);
    OutputSpec reduceOutputSpec = new OutputSpec("NullDestinationVertex",
        new OutputDescriptor(MROutputLegacy.class.getName()), 1);

    // Now run a reduce
    TaskSpec taskSpec = new TaskSpec(
View Full Code Here
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
    dag.addVertex(stage3Vertex);

    Edge edge1 = new Edge(stage1Vertex, stage2Vertex, new EdgeProperty(
        DataMovementType.SCATTER_GATHER, DataSourceType.PERSISTED,
        SchedulingType.SEQUENTIAL, new OutputDescriptor(
        OnFileSortedOutput.class.getName()), new InputDescriptor(
                ShuffledMergedInputLegacy.class.getName())));
    Edge edge2 = new Edge(stage2Vertex, stage3Vertex, new EdgeProperty(
        DataMovementType.SCATTER_GATHER, DataSourceType.PERSISTED,
        SchedulingType.SEQUENTIAL, new OutputDescriptor(
        OnFileSortedOutput.class.getName()), new InputDescriptor(
                ShuffledMergedInputLegacy.class.getName())));

    dag.addEdge(edge1);
    dag.addEdge(edge2);
View Full Code Here
122
123
124
125
126
127
128
129
130
131
132
   
   
    MapUtils.generateInputSplit(localFs, workDir, job, mapInput);
   
    InputSpec mapInputSpec = new InputSpec("NullSrcVertex",
        new InputDescriptor(MRInputLegacy.class.getName())
            .setUserPayload(MRHelpers.createMRInputPayload(job, null)),
        0);
    OutputSpec mapOutputSpec = new OutputSpec("NullDestVertex", new OutputDescriptor(LocalOnFileSorterOutput.class.getName()), 1);

    LogicalIOProcessorRuntimeTask task = MapUtils.createLogicalTask(localFs, workDir, job, 0,
View Full Code Here
189
190
191
192
193
194
195
196
197
  @Test(timeout=120000)
  public void testDelayedInit() throws Exception {
    DAG dag = SimpleVTestDAG.createDAG("DelayedInitDAG", null);
    dag.getVertex("v1").addInput("i1",
        new InputDescriptor(NoOpInput.class.getName()),
        FailingInputInitializer.class);
    runDAGAndVerify(dag, State.SUCCEEDED);
  }
View Full Code Here
479
480
481
482
483
484
485
486
487
488
489
      if (i > 0) {
        EdgeProperty edgeProperty = new EdgeProperty(
            DataMovementType.SCATTER_GATHER, DataSourceType.PERSISTED,
            SchedulingType.SEQUENTIAL,
            new OutputDescriptor(OnFileSortedOutput.class.getName()),
            new InputDescriptor(ShuffledMergedInputLegacy.class.getName()));

        Edge edge = null;
        edge = new Edge(vertices[i - 1], vertices[i], edgeProperty);
        dag.addEdge(edge);
      }
View Full Code Here
907
908
909
910
911
912
913
914
915
   * @param userPayload
   * @param initClazz class to init the input in the AM
   */
  public static void addMRInput(Vertex vertex, byte[] userPayload,
      Class<? extends TezRootInputInitializer> initClazz) {
    InputDescriptor id = new InputDescriptor(MRInputLegacy.class.getName())
        .setUserPayload(userPayload);
    vertex.addInput("MRInput", id, initClazz);
  }
View Full Code Here
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
    // Configure the Input for stage1
    Class<? extends TezRootInputInitializer> initializerClazz = generateSplitsInClient ? null
        : MRInputAMSplitGenerator.class;
    stage1Vertex.addInput("MRInput",
        new InputDescriptor(MRInputLegacy.class.getName())
            .setUserPayload(MRHelpers.createMRInputPayload(stage1Payload, null)),
        initializerClazz);

    // Setup stage2 Vertex
    Vertex stage2Vertex = new Vertex("stage2", new ProcessorDescriptor(
        FilterByWordOutputProcessor.class.getName()).setUserPayload(MRHelpers
        .createUserPayloadFromConf(stage2Conf)), 1,
        MRHelpers.getReduceResource(stage2Conf));
    stage2Vertex.setJavaOpts(MRHelpers.getReduceJavaOpts(stage2Conf)).setTaskLocalResources(commonLocalResources);
    Map<String, String> stage2Env = new HashMap<String, String>();
    MRHelpers.updateEnvironmentForMRTasks(stage2Conf, stage2Env, false);
    stage2Vertex.setTaskEnvironment(stage2Env);

    // Configure the Output for stage2
    OutputDescriptor od = new OutputDescriptor(MROutput.class.getName())
        .setUserPayload(MRHelpers.createUserPayloadFromConf(stage2Conf));
    stage2Vertex.addOutput("MROutput", od, MROutputCommitter.class);

    DAG dag = new DAG("FilterLinesByWord");
    Edge edge = new Edge(stage1Vertex, stage2Vertex, new EdgeProperty(
        DataMovementType.BROADCAST, DataSourceType.PERSISTED,
        SchedulingType.SEQUENTIAL, new OutputDescriptor(
            OnFileUnorderedKVOutput.class.getName()), new InputDescriptor(
            ShuffledUnorderedKVInput.class.getName())));
    dag.addVertex(stage1Vertex).addVertex(stage2Vertex).addEdge(edge);

    LOG.info("Submitting DAG to Tez Session");
    DAGClient dagClient = tezSession.submitDAG(dag);
View Full Code Here
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
    tokenizerVertex.setJavaOpts(MRHelpers.getMapJavaOpts(mapStageConf));
    Map<String, String> mapEnv = new HashMap<String, String>();
    MRHelpers.updateEnvironmentForMRTasks(mapStageConf, mapEnv, true);
    tokenizerVertex.setTaskEnvironment(mapEnv);
    Class<? extends TezRootInputInitializer> initializerClazz = MRInputAMSplitGenerator.class;
    InputDescriptor id = new InputDescriptor(MRInput.class.getName()).
        setUserPayload(mapInputPayload);
    tokenizerVertex.addInput("MRInput", id, initializerClazz);

    byte[] finalReducePayload = MRHelpers.createUserPayloadFromConf(finalReduceConf);
    Vertex summerVertex = new Vertex("summer",
        new ProcessorDescriptor(
            SumProcessor.class.getName()).setUserPayload(finalReducePayload),
                1, MRHelpers.getReduceResource(finalReduceConf));
    summerVertex.setJavaOpts(
        MRHelpers.getReduceJavaOpts(finalReduceConf));
    Map<String, String> reduceEnv = new HashMap<String, String>();
    MRHelpers.updateEnvironmentForMRTasks(finalReduceConf, reduceEnv, false);
    summerVertex.setTaskEnvironment(reduceEnv);
    OutputDescriptor od = new OutputDescriptor(MROutput.class.getName())
        .setUserPayload(finalReducePayload);
    summerVertex.addOutput("MROutput", od, MROutputCommitter.class);
   
    DAG dag = new DAG("WordCount");
    dag.addVertex(tokenizerVertex)
        .addVertex(summerVertex)
        .addEdge(
            new Edge(tokenizerVertex, summerVertex, new EdgeProperty(
                DataMovementType.SCATTER_GATHER, DataSourceType.PERSISTED,
                SchedulingType.SEQUENTIAL,
                new OutputDescriptor(OnFileSortedOutput.class.getName())
                        .setUserPayload(mapPayload),
                new InputDescriptor(ShuffledMergedInput.class.getName())
                        .setUserPayload(finalReducePayload))));
    return dag; 
  }
View Full Code Here
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
    mapVertex1.setJavaOpts(MRHelpers.getMapJavaOpts(mapStageConf));
    Map<String, String> mapEnv = new HashMap<String, String>();
    MRHelpers.updateEnvironmentForMRTasks(mapStageConf, mapEnv, true);
    mapVertex1.setTaskEnvironment(mapEnv);
    Class<? extends TezRootInputInitializer> initializerClazz = MRInputAMSplitGenerator.class;
    InputDescriptor id = new InputDescriptor(MRInput.class.getName()).
        setUserPayload(mapInputPayload);
    mapVertex1.addInput("MRInput", id, initializerClazz);

    Vertex mapVertex2 = new Vertex("map2", new ProcessorDescriptor(
        TokenProcessor.class.getName()),
        numMaps, MRHelpers.getMapResource(mapStageConf));
    mapVertex2.setJavaOpts(MRHelpers.getMapJavaOpts(mapStageConf));
    MRHelpers.updateEnvironmentForMRTasks(mapStageConf, mapEnv, true);
    mapVertex2.setTaskEnvironment(mapEnv);
    mapVertex2.addInput("MRInput", id, initializerClazz);

    Vertex mapVertex3 = new Vertex("map3", new ProcessorDescriptor(
        TokenProcessor.class.getName()),
        numMaps, MRHelpers.getMapResource(mapStageConf));
    mapVertex3.setJavaOpts(MRHelpers.getMapJavaOpts(mapStageConf));
    MRHelpers.updateEnvironmentForMRTasks(mapStageConf, mapEnv, true);
    mapVertex3.setTaskEnvironment(mapEnv);
    mapVertex3.addInput("MRInput", id, initializerClazz);
   
    byte[] finalReducePayload = MRHelpers.createUserPayloadFromConf(finalReduceConf);
    Vertex checkerVertex = new Vertex("checker",
        new ProcessorDescriptor(
            UnionProcessor.class.getName()).setUserPayload(finalReducePayload),
                1, MRHelpers.getReduceResource(finalReduceConf));
    checkerVertex.setJavaOpts(
        MRHelpers.getReduceJavaOpts(finalReduceConf));
    Map<String, String> reduceEnv = new HashMap<String, String>();
    MRHelpers.updateEnvironmentForMRTasks(finalReduceConf, reduceEnv, false);
    checkerVertex.setTaskEnvironment(reduceEnv);
    OutputDescriptor od = new OutputDescriptor(MROutput.class.getName())
      .setUserPayload(finalReducePayload);
    checkerVertex.addOutput("union", od, MROutputCommitter.class);

    Configuration partsConf = new Configuration(finalReduceConf);
    partsConf.set(FileOutputFormat.OUTDIR, outputPath+"-parts");
    byte[] partsPayload = MRHelpers.createUserPayloadFromConf(partsConf);
   
    DAG dag = new DAG("UnionExample");
   
    VertexGroup unionVertex = dag.createVertexGroup("union", mapVertex1, mapVertex2);
    OutputDescriptor od1 = new OutputDescriptor(MROutput.class.getName())
      .setUserPayload(partsPayload);
    Configuration allPartsConf = new Configuration(finalReduceConf);
    allPartsConf.set(FileOutputFormat.OUTDIR, outputPath+"-all-parts");
    byte[] allPartsPayload = MRHelpers.createUserPayloadFromConf(allPartsConf);
    OutputDescriptor od2 = new OutputDescriptor(MROutput.class.getName())
      .setUserPayload(allPartsPayload);
    unionVertex.addOutput("parts", od1, MROutputCommitter.class);
    checkerVertex.addOutput("all-parts", od2, MROutputCommitter.class);
   
   
    dag.addVertex(mapVertex1)
        .addVertex(mapVertex2)
        .addVertex(mapVertex3)
        .addVertex(checkerVertex)
        .addEdge(
            new Edge(mapVertex3, checkerVertex, new EdgeProperty(
                DataMovementType.SCATTER_GATHER, DataSourceType.PERSISTED,
                SchedulingType.SEQUENTIAL,
                new OutputDescriptor(OnFileSortedOutput.class.getName())
                        .setUserPayload(mapPayload),
                new InputDescriptor(ShuffledMergedInput.class.getName())
                        .setUserPayload(finalReducePayload))))
        .addEdge(
            new GroupInputEdge(unionVertex, checkerVertex, new EdgeProperty(
                DataMovementType.SCATTER_GATHER, DataSourceType.PERSISTED,
                SchedulingType.SEQUENTIAL,
                new OutputDescriptor(OnFileSortedOutput.class.getName())
                    .setUserPayload(mapPayload),
                new InputDescriptor(ShuffledMergedInput.class.getName())
                    .setUserPayload(finalReducePayload)),
                new InputDescriptor(
                    ConcatenatedMergedKeyValuesInput.class.getName())));
    return dag; 
  }
View Full Code Here
TOP

Related Classes of org.apache.tez.dag.api.InputDescriptor

  • cascading.flow.tez.Hadoop2TezFlowStep
  • org.apache.hadoop.hive.ql.exec.tez.DagUtils
  • org.apache.pig.backend.hadoop.executionengine.tez.TezDagBuilder
  • org.apache.tez.dag.app.dag.impl.TestDAGImpl
  • org.apache.tez.dag.app.dag.impl.TestRootInputVertexManager
  • org.apache.tez.dag.app.dag.impl.TestVertexScheduler
  • org.apache.tez.dag.app.dag.impl.VertexImpl
  • org.apache.tez.dag.app.dag.impl.VertexImpl$InitTransition
  • org.apache.tez.dag.app.rm.TestContainerReuse
  • org.apache.tez.dag.library.vertexmanager.TestShuffleVertexManager
    • Copyright © 2018 www.massapicom. All rights reserved.
    All source code are property of their respective owners. Java is a trademark of Sun Microsystems, Inc and owned by ORACLE Inc. Contact coftware#gmail.com.