Package eu.stratosphere.nephele.jobgraph

Examples of eu.stratosphere.nephele.jobgraph.JobTaskVertex

187
188
189
190
191
192
193
194
195
196
197
198
199
200
 
  private static JobTaskVertex createMapper(JobGraph jobGraph, int numSubTasks, TypeSerializerFactory<?> inputSerializer,
      TypeSerializerFactory<?> broadcastVarSerializer, TypeSerializerFactory<?> outputSerializer,
      TypeComparatorFactory<?> outputComparator)
  {
    JobTaskVertex mapper = JobGraphUtils.createTask(IterationIntermediatePactTask.class,
      "Map (Select nearest center)", jobGraph, numSubTasks, numSubTasks);
   
    TaskConfig intermediateConfig = new TaskConfig(mapper.getConfiguration());
    intermediateConfig.setIterationId(ITERATION_ID);
   
    intermediateConfig.setDriver(CollectorMapDriver.class);
    intermediateConfig.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
    intermediateConfig.addInputToGroup(0);
View Full Code Here
217
218
219
220
221
222
223
224
225
226
227
228
229
230
  private static JobTaskVertex createReducer(JobGraph jobGraph, int numSubTasks, TypeSerializerFactory<?> inputSerializer,
      TypeComparatorFactory<?> inputComparator, TypeSerializerFactory<?> outputSerializer)
  {
    // ---------------- the tail (co group) --------------------
   
    JobTaskVertex tail = JobGraphUtils.createTask(IterationTailPactTask.class, "Reduce / Iteration Tail", jobGraph,
      numSubTasks, numSubTasks);
   
    TaskConfig tailConfig = new TaskConfig(tail.getConfiguration());
    tailConfig.setIterationId(ITERATION_ID);
    tailConfig.setIsWorksetUpdate();
   
    // inputs and driver
    tailConfig.setDriver(GroupReduceDriver.class);
View Full Code Here
272
273
274
275
276
277
278
279
280
281
282
283
284
285
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
    // -- vertices ---------------------------------------------------------------------------------------------
    JobInputVertex points = createPointsInput(jobGraph, pointsPath, numSubTasks, serializer);
    JobInputVertex centers = createCentersInput(jobGraph, centersPath, numSubTasks, serializer);
   
    JobTaskVertex head = createIterationHead(jobGraph, numSubTasks, serializer);
    JobTaskVertex mapper = createMapper(jobGraph, numSubTasks, serializer, serializer, serializer, int0Comparator);
   
    JobTaskVertex reducer = createReducer(jobGraph, numSubTasks, serializer, int0Comparator, serializer);
   
    JobOutputVertex fakeTailOutput = JobGraphUtils.createFakeOutput(jobGraph, "FakeTailOutput", numSubTasks, numSubTasks);
   
    JobOutputVertex sync = createSync(jobGraph, numIterations, numSubTasks);
   
    JobOutputVertex output = createOutput(jobGraph, resultPath, numSubTasks, serializer);

    // -- edges ------------------------------------------------------------------------------------------------
    JobGraphUtils.connect(points, mapper, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    JobGraphUtils.connect(centers, head, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    JobGraphUtils.connect(head, mapper, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    new TaskConfig(mapper.getConfiguration()).setBroadcastGateIterativeWithNumberOfEventsUntilInterrupt(0, numSubTasks);
    new TaskConfig(mapper.getConfiguration()).setInputCached(0, true);
    new TaskConfig(mapper.getConfiguration()).setInputMaterializationMemory(0, MEMORY_PER_CONSUMER * JobGraphUtils.MEGABYTE);

    JobGraphUtils.connect(mapper, reducer, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    new TaskConfig(reducer.getConfiguration()).setGateIterativeWithNumberOfEventsUntilInterrupt(0, numSubTasks);
   
    JobGraphUtils.connect(reducer, fakeTailOutput, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    JobGraphUtils.connect(head, output, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    JobGraphUtils.connect(head, sync, ChannelType.NETWORK, DistributionPattern.BIPARTITE);

    // -- instance sharing -------------------------------------------------------------------------------------
    points.setVertexToShareInstancesWith(output);
    centers.setVertexToShareInstancesWith(output);
    head.setVertexToShareInstancesWith(output);
    mapper.setVertexToShareInstancesWith(output);
    reducer.setVertexToShareInstancesWith(output);
    fakeTailOutput.setVertexToShareInstancesWith(output);
    sync.setVertexToShareInstancesWith(output);

    return jobGraph;
  }
View Full Code Here
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
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
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
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
    adjacencyListInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    adjacencyListInputConfig.setOutputSerializer(vertexWithAdjacencyListSerializer);
    adjacencyListInputConfig.setOutputComparator(vertexWithAdjacencyListComparator, 0);

    // --------------- the head ---------------------
    JobTaskVertex head = JobGraphUtils.createTask(IterationHeadPactTask.class, "IterationHead", jobGraph,
      degreeOfParallelism, numSubTasksPerInstance);
    TaskConfig headConfig = new TaskConfig(head.getConfiguration());
    headConfig.setIterationId(ITERATION_ID);
   
    // initial input / partial solution
    headConfig.addInputToGroup(0);
    headConfig.setIterationHeadPartialSolutionOrWorksetInputIndex(0);
    headConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    headConfig.setInputComparator(vertexWithRankAndDanglingComparator, 0);
    headConfig.setInputLocalStrategy(0, LocalStrategy.SORT);
    headConfig.setMemoryInput(0, minorConsumer * JobGraphUtils.MEGABYTE);
    headConfig.setFilehandlesInput(0, NUM_FILE_HANDLES_PER_SORT);
    headConfig.setSpillingThresholdInput(0, SORT_SPILL_THRESHOLD);
   
    // back channel / iterations
    headConfig.setBackChannelMemory(minorConsumer * JobGraphUtils.MEGABYTE);
   
    // output into iteration
    headConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
    headConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    headConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
   
    // final output
    TaskConfig headFinalOutConfig = new TaskConfig(new Configuration());
    headFinalOutConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
    headFinalOutConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    headConfig.setIterationHeadFinalOutputConfig(headFinalOutConfig);
   
    // the sync
    headConfig.setIterationHeadIndexOfSyncOutput(3);
    headConfig.setNumberOfIterations(numIterations);
   
    // the driver
    headConfig.setDriver(CollectorMapDriver.class);
    headConfig.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
    headConfig.setStubWrapper(new UserCodeClassWrapper<CustomCompensatingMap>(CustomCompensatingMap.class));
    headConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));
    headConfig.setStubParameter("compensation.failingWorker", failingWorkers);
    headConfig.setStubParameter("compensation.failingIteration", String.valueOf(failingIteration));
    headConfig.setStubParameter("compensation.messageLoss", String.valueOf(messageLoss));
    headConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, PageRankStatsAggregator.class);

    // --------------- the join ---------------------
   
    JobTaskVertex intermediate = JobGraphUtils.createTask(IterationIntermediatePactTask.class,
      "IterationIntermediate", jobGraph, degreeOfParallelism, numSubTasksPerInstance);
    TaskConfig intermediateConfig = new TaskConfig(intermediate.getConfiguration());
    intermediateConfig.setIterationId(ITERATION_ID);
//    intermediateConfig.setDriver(RepeatableHashjoinMatchDriverWithCachedBuildside.class);
    intermediateConfig.setDriver(BuildSecondCachedMatchDriver.class);
    intermediateConfig.setDriverStrategy(DriverStrategy.HYBRIDHASH_BUILD_SECOND);
    intermediateConfig.setMemoryDriver(matchMemory * JobGraphUtils.MEGABYTE);
    intermediateConfig.addInputToGroup(0);
    intermediateConfig.addInputToGroup(1);
    intermediateConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    intermediateConfig.setInputSerializer(vertexWithAdjacencyListSerializer, 1);
    intermediateConfig.setDriverComparator(vertexWithRankAndDanglingComparator, 0);
    intermediateConfig.setDriverComparator(vertexWithAdjacencyListComparator, 1);
    intermediateConfig.setDriverPairComparator(matchComparator);
   
    intermediateConfig.setOutputSerializer(vertexWithRankSerializer);
    intermediateConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    intermediateConfig.setOutputComparator(vertexWithRankComparator, 0);
   
    intermediateConfig.setStubWrapper(new UserCodeClassWrapper<CustomCompensatableDotProductMatch>(CustomCompensatableDotProductMatch.class));
    intermediateConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));
    intermediateConfig.setStubParameter("compensation.failingWorker", failingWorkers);
    intermediateConfig.setStubParameter("compensation.failingIteration", String.valueOf(failingIteration));
    intermediateConfig.setStubParameter("compensation.messageLoss", String.valueOf(messageLoss));

    // ---------------- the tail (co group) --------------------
   
    JobTaskVertex tail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationTail", jobGraph,
      degreeOfParallelism, numSubTasksPerInstance);
    TaskConfig tailConfig = new TaskConfig(tail.getConfiguration());
    tailConfig.setIterationId(ITERATION_ID);
        tailConfig.setIsWorksetUpdate();
    // TODO we need to combine!
   
    // inputs and driver
    tailConfig.setDriver(CoGroupDriver.class);
    tailConfig.setDriverStrategy(DriverStrategy.CO_GROUP);
    tailConfig.addInputToGroup(0);
    tailConfig.addInputToGroup(1);
    tailConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    tailConfig.setInputSerializer(vertexWithRankSerializer, 1);
    tailConfig.setDriverComparator(vertexWithRankAndDanglingComparator, 0);
    tailConfig.setDriverComparator(vertexWithRankComparator, 1);
    tailConfig.setDriverPairComparator(coGroupComparator);
    tailConfig.setInputAsynchronouslyMaterialized(0, true);
    tailConfig.setInputMaterializationMemory(0, minorConsumer * JobGraphUtils.MEGABYTE);
    tailConfig.setInputLocalStrategy(1, LocalStrategy.SORT);
    tailConfig.setInputComparator(vertexWithRankComparator, 1);
    tailConfig.setMemoryInput(1, coGroupSortMemory * JobGraphUtils.MEGABYTE);
    tailConfig.setFilehandlesInput(1, NUM_FILE_HANDLES_PER_SORT);
    tailConfig.setSpillingThresholdInput(1, SORT_SPILL_THRESHOLD);
   
    // output
    tailConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    tailConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
   
    // the stub
    tailConfig.setStubWrapper(new UserCodeClassWrapper<CustomCompensatableDotProductCoGroup>(CustomCompensatableDotProductCoGroup.class));
    tailConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));
    tailConfig.setStubParameter("pageRank.numDanglingVertices", String.valueOf(numDanglingVertices));
    tailConfig.setStubParameter("compensation.failingWorker", failingWorkers);
    tailConfig.setStubParameter("compensation.failingIteration", String.valueOf(failingIteration));
    tailConfig.setStubParameter("compensation.messageLoss", String.valueOf(messageLoss));
   
    // --------------- the output ---------------------

    JobOutputVertex output = JobGraphUtils.createFileOutput(jobGraph, "FinalOutput", degreeOfParallelism,
      numSubTasksPerInstance);
    TaskConfig outputConfig = new TaskConfig(output.getConfiguration());
    outputConfig.addInputToGroup(0);
    outputConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    outputConfig.setStubWrapper(new UserCodeClassWrapper<CustomPageWithRankOutFormat>(CustomPageWithRankOutFormat.class));
    outputConfig.setStubParameter(FileOutputFormat.FILE_PARAMETER_KEY, outputPath);
   
    // --------------- the auxiliaries ---------------------
   
    JobOutputVertex fakeTailOutput = JobGraphUtils.createFakeOutput(jobGraph, "FakeTailOutput",
      degreeOfParallelism, numSubTasksPerInstance);

    JobOutputVertex sync = JobGraphUtils.createSync(jobGraph, degreeOfParallelism);
    TaskConfig syncConfig = new TaskConfig(sync.getConfiguration());
    syncConfig.setNumberOfIterations(numIterations);
    syncConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, PageRankStatsAggregator.class);
    syncConfig.setConvergenceCriterion(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, DiffL1NormConvergenceCriterion.class);
    syncConfig.setIterationId(ITERATION_ID);
   
    // --------------- the wiring ---------------------

    JobGraphUtils.connect(pageWithRankInput, head, ChannelType.NETWORK, DistributionPattern.BIPARTITE);

    JobGraphUtils.connect(head, intermediate, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    intermediateConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);
   
    JobGraphUtils.connect(adjacencyListInput, intermediate, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
   
    JobGraphUtils.connect(head, tail, ChannelType.NETWORK, DistributionPattern.POINTWISE);
    JobGraphUtils.connect(intermediate, tail, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    tailConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);
    tailConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(1, degreeOfParallelism);

    JobGraphUtils.connect(head, output, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    JobGraphUtils.connect(tail, fakeTailOutput, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);

    JobGraphUtils.connect(head, sync, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    fakeTailOutput.setVertexToShareInstancesWith(tail);
    tail.setVertexToShareInstancesWith(head);
    pageWithRankInput.setVertexToShareInstancesWith(head);
    adjacencyListInput.setVertexToShareInstancesWith(head);
    intermediate.setVertexToShareInstancesWith(head);
    output.setVertexToShareInstancesWith(head);
    sync.setVertexToShareInstancesWith(head);
View Full Code Here
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
      final JobFileInputVertex i1 = new JobFileInputVertex("Input 1", jg);
      i1.setFileInputClass(FileLineReader.class);
      i1.setFilePath(new Path(inputFile.toURI()));

      // task vertex
      final JobTaskVertex t1 = new JobTaskVertex("Task 1", jg);
      t1.setTaskClass(ForwardTask1Input1Output.class);

      // output vertex
      final JobFileOutputVertex o1 = new JobFileOutputVertex("Output 1", jg);
      o1.setFileOutputClass(FileLineWriter.class);
      o1.setFilePath(new Path(new File(ServerTestUtils.getRandomFilename()).toURI()));

      o1.setVertexToShareInstancesWith(i1);
      i1.setVertexToShareInstancesWith(t1);

      // connect vertices
      i1.connectTo(t1);
      t1.connectTo(o1);

      LibraryCacheManager.register(jobID, new String[0]);

      final ExecutionGraph eg = new ExecutionGraph(jg, INSTANCE_MANAGER);
View Full Code Here
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
      final JobFileInputVertex i1 = new JobFileInputVertex("Input 1", jg);
      i1.setFileInputClass(FileLineReader.class);
      i1.setFilePath(new Path(inputFile.toURI()));

      // task vertex
      final JobTaskVertex t1 = new JobTaskVertex("Task 1", jg);
      t1.setTaskClass(ForwardTask1Input1Output.class);

      // output vertex
      final JobFileOutputVertex o1 = new JobFileOutputVertex("Output 1", jg);
      o1.setFileOutputClass(FileLineWriter.class);
      o1.setFilePath(new Path(new File(ServerTestUtils.getRandomFilename()).toURI()));

      // connect vertices
      i1.connectTo(t1, ChannelType.IN_MEMORY);
      t1.connectTo(o1, ChannelType.IN_MEMORY);

      LibraryCacheManager.register(jobID, new String[0]);

      // now convert job graph to execution graph
      final ExecutionGraph eg = new ExecutionGraph(jg, INSTANCE_MANAGER);
View Full Code Here
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
      i2.setFileInputClass(FileLineReader.class);
      i2.setFilePath(new Path(inputFile2.toURI()));
      i2.setNumberOfSubtasks(2);

      // task vertex
      final JobTaskVertex t1 = new JobTaskVertex("Task 1", jg);
      t1.setTaskClass(ForwardTask1Input1Output.class);
      t1.setNumberOfSubtasks(2);
      final JobTaskVertex t2 = new JobTaskVertex("Task 2", jg);
      t2.setTaskClass(ForwardTask1Input1Output.class);
      t2.setNumberOfSubtasks(2);
      final JobTaskVertex t3 = new JobTaskVertex("Task 3", jg);
      t3.setTaskClass(ForwardTask2Inputs1Output.class);
      t3.setNumberOfSubtasks(2);

     
      // output vertex
      final JobFileOutputVertex o1 = new JobFileOutputVertex("Output 1", jg);
      o1.setFileOutputClass(FileLineWriter.class);
      o1.setFilePath(new Path(outputFile.toURI()));
      o1.setNumberOfSubtasks(2);
      i1.setVertexToShareInstancesWith(t1);
      t1.setVertexToShareInstancesWith(t3);
      i2.setVertexToShareInstancesWith(t2);
      t2.setVertexToShareInstancesWith(t3);
      t3.setVertexToShareInstancesWith(o1);

      // connect vertices
      i1.connectTo(t1);
      i2.connectTo(t2);
      t1.connectTo(t3);
      t2.connectTo(t3);
      t3.connectTo(o1);

      LibraryCacheManager.register(jobID, new String[0]);

      final ExecutionGraph eg = new ExecutionGraph(jg, INSTANCE_MANAGER);
View Full Code Here
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
      i2.setFileInputClass(FileLineReader.class);
      i2.setFilePath(new Path(inputFile2.toURI()));
      i2.setNumberOfSubtasks(4);
      i2.setNumberOfSubtasksPerInstance(2);
      // task vertex
      final JobTaskVertex t1 = new JobTaskVertex("Task 1", jg);
      t1.setTaskClass(ForwardTask1Input1Output.class);
      t1.setNumberOfSubtasks(4);
      t1.setNumberOfSubtasksPerInstance(2);
      final JobTaskVertex t2 = new JobTaskVertex("Task 2", jg);
      t2.setTaskClass(ForwardTask1Input1Output.class);
      t2.setNumberOfSubtasks(4);
      t2.setNumberOfSubtasksPerInstance(2);
      final JobTaskVertex t3 = new JobTaskVertex("Task 3", jg);
      t3.setTaskClass(ForwardTask2Inputs1Output.class);
      t3.setNumberOfSubtasks(8);
      t3.setNumberOfSubtasksPerInstance(4);
      final JobTaskVertex t4 = new JobTaskVertex("Task 4", jg);
      t4.setTaskClass(ForwardTask1Input2Outputs.class);
      t4.setNumberOfSubtasks(8);
      t4.setNumberOfSubtasksPerInstance(4);
      // output vertex
      final JobFileOutputVertex o1 = new JobFileOutputVertex("Output 1", jg);
      o1.setFileOutputClass(FileLineWriter.class);
      o1.setFilePath(new Path(outputFile1.toURI()));
      o1.setNumberOfSubtasks(4);
      o1.setNumberOfSubtasksPerInstance(2);
      final JobFileOutputVertex o2 = new JobFileOutputVertex("Output 2", jg);
      o2.setFileOutputClass(FileLineWriter.class);
      o2.setFilePath(new Path(outputFile2.toURI()));
      o2.setNumberOfSubtasks(4);
      o2.setNumberOfSubtasksPerInstance(2);
      o1.setVertexToShareInstancesWith(o2);

      // connect vertices
      i1.connectTo(t1, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
      i2.connectTo(t2, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
      t1.connectTo(t3, ChannelType.NETWORK);
      t2.connectTo(t3, ChannelType.NETWORK);
      t3.connectTo(t4, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
      t4.connectTo(o1, ChannelType.NETWORK);
      t4.connectTo(o2, ChannelType.NETWORK);

      LibraryCacheManager.register(jobID, new String[0]);

      // now convert job graph to execution graph
      final ExecutionGraph eg = new ExecutionGraph(jg, INSTANCE_MANAGER);
View Full Code Here
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
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
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
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
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
    adjacencyListInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    adjacencyListInputConfig.setOutputSerializer(vertexWithAdjacencyListSerializer);
    adjacencyListInputConfig.setOutputComparator(vertexWithAdjacencyListComparator, 0);

    // --------------- the head ---------------------
    JobTaskVertex head = JobGraphUtils.createTask(IterationHeadPactTask.class, "IterationHead", jobGraph,
      degreeOfParallelism, numSubTasksPerInstance);
    TaskConfig headConfig = new TaskConfig(head.getConfiguration());
    headConfig.setIterationId(ITERATION_ID);
   
    // initial input / partial solution
    headConfig.addInputToGroup(0);
    headConfig.setIterationHeadPartialSolutionOrWorksetInputIndex(0);
    headConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    headConfig.setInputComparator(vertexWithRankAndDanglingComparator, 0);
    headConfig.setInputLocalStrategy(0, LocalStrategy.SORT);
    headConfig.setMemoryInput(0, minorConsumer * JobGraphUtils.MEGABYTE);
    headConfig.setFilehandlesInput(0, NUM_FILE_HANDLES_PER_SORT);
    headConfig.setSpillingThresholdInput(0, SORT_SPILL_THRESHOLD);
   
    // back channel / iterations
    headConfig.setBackChannelMemory(minorConsumer * JobGraphUtils.MEGABYTE);
   
    // output into iteration
    headConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
    headConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    headConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
   
    // final output
    TaskConfig headFinalOutConfig = new TaskConfig(new Configuration());
    headFinalOutConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
    headFinalOutConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    headConfig.setIterationHeadFinalOutputConfig(headFinalOutConfig);
   
    // the sync
    headConfig.setIterationHeadIndexOfSyncOutput(3);
    headConfig.setNumberOfIterations(numIterations);
   
    // the driver
    headConfig.setDriver(CollectorMapDriver.class);
    headConfig.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
    headConfig.setStubWrapper(new UserCodeClassWrapper<CustomCompensatingMap>(CustomCompensatingMap.class));
    headConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));
    headConfig.setStubParameter("compensation.failingWorker", failingWorkers);
    headConfig.setStubParameter("compensation.failingIteration", String.valueOf(failingIteration));
    headConfig.setStubParameter("compensation.messageLoss", String.valueOf(messageLoss));
    headConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, PageRankStatsAggregator.class);

    // --------------- the join ---------------------
   
    JobTaskVertex intermediate = JobGraphUtils.createTask(IterationIntermediatePactTask.class,
      "IterationIntermediate", jobGraph, degreeOfParallelism, numSubTasksPerInstance);
    TaskConfig intermediateConfig = new TaskConfig(intermediate.getConfiguration());
    intermediateConfig.setIterationId(ITERATION_ID);
//    intermediateConfig.setDriver(RepeatableHashjoinMatchDriverWithCachedBuildside.class);
    intermediateConfig.setDriver(BuildSecondCachedMatchDriver.class);
    intermediateConfig.setDriverStrategy(DriverStrategy.HYBRIDHASH_BUILD_SECOND);
    intermediateConfig.setMemoryDriver(matchMemory * JobGraphUtils.MEGABYTE);
    intermediateConfig.addInputToGroup(0);
    intermediateConfig.addInputToGroup(1);
    intermediateConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    intermediateConfig.setInputSerializer(vertexWithAdjacencyListSerializer, 1);
    intermediateConfig.setDriverComparator(vertexWithRankAndDanglingComparator, 0);
    intermediateConfig.setDriverComparator(vertexWithAdjacencyListComparator, 1);
    intermediateConfig.setDriverPairComparator(matchComparator);
   
    intermediateConfig.setOutputSerializer(vertexWithRankSerializer);
    intermediateConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
   
    intermediateConfig.setStubWrapper(new UserCodeClassWrapper<CustomCompensatableDotProductMatch>(CustomCompensatableDotProductMatch.class));
    intermediateConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));
    intermediateConfig.setStubParameter("compensation.failingWorker", failingWorkers);
    intermediateConfig.setStubParameter("compensation.failingIteration", String.valueOf(failingIteration));
    intermediateConfig.setStubParameter("compensation.messageLoss", String.valueOf(messageLoss));
   
    // the combiner and the output
    TaskConfig combinerConfig = new TaskConfig(new Configuration());
    combinerConfig.addInputToGroup(0);
    combinerConfig.setInputSerializer(vertexWithRankSerializer, 0);
    combinerConfig.setDriverStrategy(DriverStrategy.SORTED_GROUP_COMBINE);
    combinerConfig.setDriverComparator(vertexWithRankComparator, 0);
    combinerConfig.setMemoryDriver(coGroupSortMemory * JobGraphUtils.MEGABYTE);
    combinerConfig.setOutputSerializer(vertexWithRankSerializer);
    combinerConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    combinerConfig.setOutputComparator(vertexWithRankComparator, 0);
    combinerConfig.setStubWrapper(new UserCodeClassWrapper<CustomRankCombiner>(CustomRankCombiner.class));
    intermediateConfig.addChainedTask(SynchronousChainedCombineDriver.class, combinerConfig, "Combiner");

    // ---------------- the tail (co group) --------------------
   
    JobTaskVertex tail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationTail", jobGraph,
      degreeOfParallelism, numSubTasksPerInstance);
    TaskConfig tailConfig = new TaskConfig(tail.getConfiguration());
    tailConfig.setIterationId(ITERATION_ID);
        tailConfig.setIsWorksetUpdate();
   
    // inputs and driver
    tailConfig.setDriver(CoGroupDriver.class);
    tailConfig.setDriverStrategy(DriverStrategy.CO_GROUP);
    tailConfig.addInputToGroup(0);
    tailConfig.addInputToGroup(1);
    tailConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    tailConfig.setInputSerializer(vertexWithRankSerializer, 1);
    tailConfig.setDriverComparator(vertexWithRankAndDanglingComparator, 0);
    tailConfig.setDriverComparator(vertexWithRankComparator, 1);
    tailConfig.setDriverPairComparator(coGroupComparator);
    tailConfig.setInputAsynchronouslyMaterialized(0, true);
    tailConfig.setInputMaterializationMemory(0, minorConsumer * JobGraphUtils.MEGABYTE);
    tailConfig.setInputLocalStrategy(1, LocalStrategy.SORT);
    tailConfig.setInputComparator(vertexWithRankComparator, 1);
    tailConfig.setMemoryInput(1, coGroupSortMemory * JobGraphUtils.MEGABYTE);
    tailConfig.setFilehandlesInput(1, NUM_FILE_HANDLES_PER_SORT);
    tailConfig.setSpillingThresholdInput(1, SORT_SPILL_THRESHOLD);
    tailConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, PageRankStatsAggregator.class);
   
    // output
    tailConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    tailConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
   
    // the stub
    tailConfig.setStubWrapper(new UserCodeClassWrapper<CustomCompensatableDotProductCoGroup>(CustomCompensatableDotProductCoGroup.class));
    tailConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));
    tailConfig.setStubParameter("pageRank.numDanglingVertices", String.valueOf(numDanglingVertices));
    tailConfig.setStubParameter("compensation.failingWorker", failingWorkers);
    tailConfig.setStubParameter("compensation.failingIteration", String.valueOf(failingIteration));
    tailConfig.setStubParameter("compensation.messageLoss", String.valueOf(messageLoss));
   
    // --------------- the output ---------------------

    JobOutputVertex output = JobGraphUtils.createFileOutput(jobGraph, "FinalOutput", degreeOfParallelism,
      numSubTasksPerInstance);
    TaskConfig outputConfig = new TaskConfig(output.getConfiguration());
    outputConfig.addInputToGroup(0);
    outputConfig.setInputSerializer(vertexWithRankAndDanglingSerializer, 0);
    outputConfig.setStubWrapper(new UserCodeClassWrapper<CustomPageWithRankOutFormat>(CustomPageWithRankOutFormat.class));
    outputConfig.setStubParameter(FileOutputFormat.FILE_PARAMETER_KEY, outputPath);
   
    // --------------- the auxiliaries ---------------------
   
    JobOutputVertex fakeTailOutput = JobGraphUtils.createFakeOutput(jobGraph, "FakeTailOutput",
      degreeOfParallelism, numSubTasksPerInstance);

    JobOutputVertex sync = JobGraphUtils.createSync(jobGraph, degreeOfParallelism);
    TaskConfig syncConfig = new TaskConfig(sync.getConfiguration());
    syncConfig.setNumberOfIterations(numIterations);
    syncConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, PageRankStatsAggregator.class);
    syncConfig.setConvergenceCriterion(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, DiffL1NormConvergenceCriterion.class);
    syncConfig.setIterationId(ITERATION_ID);
   
    // --------------- the wiring ---------------------

    JobGraphUtils.connect(pageWithRankInput, head, ChannelType.NETWORK, DistributionPattern.BIPARTITE);

    JobGraphUtils.connect(head, intermediate, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    intermediateConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);
   
    JobGraphUtils.connect(adjacencyListInput, intermediate, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
   
    JobGraphUtils.connect(head, tail, ChannelType.NETWORK, DistributionPattern.POINTWISE);
    JobGraphUtils.connect(intermediate, tail, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    tailConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);
    tailConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(1, degreeOfParallelism);

    JobGraphUtils.connect(head, output, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    JobGraphUtils.connect(tail, fakeTailOutput, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);

    JobGraphUtils.connect(head, sync, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    fakeTailOutput.setVertexToShareInstancesWith(tail);
    tail.setVertexToShareInstancesWith(head);
    pageWithRankInput.setVertexToShareInstancesWith(head);
    adjacencyListInput.setVertexToShareInstancesWith(head);
    intermediate.setVertexToShareInstancesWith(head);
    output.setVertexToShareInstancesWith(head);
    sync.setVertexToShareInstancesWith(head);
View Full Code Here
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
    // input
    JobInputVertex vertices = createVerticesInput(jobGraph, verticesPath, numSubTasks, serializer, comparator);
    JobInputVertex edges = createEdgesInput(jobGraph, edgesPath, numSubTasks, serializer, comparator);

    // head
    JobTaskVertex head = createIterationHead(jobGraph, numSubTasks, serializer, comparator, pairComparator);
    TaskConfig headConfig = new TaskConfig(head.getConfiguration());
    headConfig.setWaitForSolutionSetUpdate();

    // intermediate
    JobTaskVertex intermediate = createIterationIntermediate(jobGraph, numSubTasks, serializer, comparator);
    TaskConfig intermediateConfig = new TaskConfig(intermediate.getConfiguration());

    // output and auxiliaries
    JobOutputVertex output = createOutput(jobGraph, resultPath, numSubTasks, serializer);
    JobOutputVertex ssFakeTail = createFakeTail(jobGraph, numSubTasks);
    JobOutputVertex wsFakeTail = createFakeTail(jobGraph, numSubTasks);
    JobOutputVertex sync = createSync(jobGraph, numSubTasks, maxIterations);

    // ------------------ the intermediate (ss join) ----------------------
    JobTaskVertex ssJoinIntermediate = JobGraphUtils.createTask(IterationIntermediatePactTask.class,
      "Solution Set Join", jobGraph, numSubTasks, numSubTasks);
    TaskConfig ssJoinIntermediateConfig = new TaskConfig(ssJoinIntermediate.getConfiguration());
    {
      ssJoinIntermediateConfig.setIterationId(ITERATION_ID);

      // inputs
      ssJoinIntermediateConfig.addInputToGroup(0);
      ssJoinIntermediateConfig.setInputSerializer(serializer, 0);

      // output
      ssJoinIntermediateConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      ssJoinIntermediateConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      ssJoinIntermediateConfig.setOutputComparator(comparator, 0);
      ssJoinIntermediateConfig.setOutputComparator(comparator, 1);

      ssJoinIntermediateConfig.setOutputSerializer(serializer);

      // driver
      ssJoinIntermediateConfig.setDriver(JoinWithSolutionSetSecondDriver.class);
      ssJoinIntermediateConfig.setDriverStrategy(DriverStrategy.HYBRIDHASH_BUILD_SECOND);
      ssJoinIntermediateConfig.setDriverComparator(comparator, 0);
      ssJoinIntermediateConfig.setDriverPairComparator(pairComparator);
     
      ssJoinIntermediateConfig.setStubWrapper(
        new UserCodeClassWrapper<UpdateComponentIdMatch>(UpdateComponentIdMatch.class));
    }

    // -------------------------- ss tail --------------------------------
    JobTaskVertex ssTail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationSolutionSetTail",
      jobGraph, numSubTasks, numSubTasks);
    TaskConfig ssTailConfig = new TaskConfig(ssTail.getConfiguration());
    {
      ssTailConfig.setIterationId(ITERATION_ID);
      ssTailConfig.setIsSolutionSetUpdate();
      ssTailConfig.setIsWorksetIteration();

      // inputs and driver
      ssTailConfig.addInputToGroup(0);
      ssTailConfig.setInputSerializer(serializer, 0);
      ssTailConfig.setInputAsynchronouslyMaterialized(0, true);
      ssTailConfig.setInputMaterializationMemory(0, MEM_PER_CONSUMER * JobGraphUtils.MEGABYTE);

      // output
      ssTailConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      ssTailConfig.setOutputSerializer(serializer);

      // the driver
      ssTailConfig.setDriver(CollectorMapDriver.class);
      ssTailConfig.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
      ssTailConfig.setStubWrapper(new UserCodeClassWrapper<DummyMapper>(DummyMapper.class));
    }

    // -------------------------- ws tail --------------------------------
    JobTaskVertex wsTail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationWorksetTail",
      jobGraph, numSubTasks, numSubTasks);
    TaskConfig wsTailConfig = new TaskConfig(wsTail.getConfiguration());
    {
      wsTailConfig.setIterationId(ITERATION_ID);
      wsTailConfig.setIsWorksetIteration();
      wsTailConfig.setIsWorksetUpdate();

      // inputs and driver
      wsTailConfig.addInputToGroup(0);
      wsTailConfig.setInputSerializer(serializer, 0);

      // output
      wsTailConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      wsTailConfig.setOutputSerializer(serializer);

      // the driver
      wsTailConfig.setDriver(CollectorMapDriver.class);
      wsTailConfig.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
      wsTailConfig.setStubWrapper(new UserCodeClassWrapper<DummyMapper>(DummyMapper.class));
    }

    // --------------- the wiring ---------------------

    JobGraphUtils.connect(vertices, head, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    JobGraphUtils.connect(edges, head, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    JobGraphUtils.connect(vertices, head, ChannelType.NETWORK, DistributionPattern.BIPARTITE);

    JobGraphUtils.connect(head, intermediate, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    intermediateConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, numSubTasks);

    JobGraphUtils.connect(intermediate, ssJoinIntermediate, ChannelType.NETWORK, DistributionPattern.POINTWISE);
    ssJoinIntermediateConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);

    JobGraphUtils.connect(ssJoinIntermediate, ssTail, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    ssTailConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);

    JobGraphUtils.connect(ssJoinIntermediate, wsTail, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    wsTailConfig.setGateIterativeWithNumberOfEventsUntilInterrupt(0, 1);

    JobGraphUtils.connect(head, output, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);

    JobGraphUtils.connect(ssTail, ssFakeTail, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);
    JobGraphUtils.connect(wsTail, wsFakeTail, ChannelType.IN_MEMORY, DistributionPattern.POINTWISE);

    JobGraphUtils.connect(head, sync, ChannelType.NETWORK, DistributionPattern.POINTWISE);

    vertices.setVertexToShareInstancesWith(head);
    edges.setVertexToShareInstancesWith(head);

    intermediate.setVertexToShareInstancesWith(head);

    ssJoinIntermediate.setVertexToShareInstancesWith(head);
    wsTail.setVertexToShareInstancesWith(head);

    output.setVertexToShareInstancesWith(head);
    sync.setVertexToShareInstancesWith(head);

    ssTail.setVertexToShareInstancesWith(wsTail);
View Full Code Here
TOP

Related Classes of eu.stratosphere.nephele.jobgraph.JobTaskVertex

  • eu.stratosphere.compiler.plantranslate.NepheleJobGraphGenerator
  • eu.stratosphere.nephele.executiongraph.ExecutionGraphTest
  • eu.stratosphere.nephele.jobmanager.JobManagerITCase
  • eu.stratosphere.test.broadcastvars.BroadcastVarsNepheleITCase
  • eu.stratosphere.test.broadcastvars.KMeansIterativeNepheleITCase
  • eu.stratosphere.test.iterative.nephele.ConnectedComponentsNepheleITCase
  • eu.stratosphere.test.iterative.nephele.customdanglingpagerank.CustomCompensatableDanglingPageRank
  • eu.stratosphere.test.iterative.nephele.customdanglingpagerank.CustomCompensatableDanglingPageRankWithCombiner
  • eu.stratosphere.test.iterative.nephele.danglingpagerank.CompensatableDanglingPageRank
  • eu.stratosphere.test.iterative.nephele.IterationWithChainingNepheleITCase
    • 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.