Package org.apache.flink.runtime.jobgraph

Examples of org.apache.flink.runtime.jobgraph.InputFormatVertex

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
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
    JobGraph jobGraph = new JobGraph("CompensatableDanglingPageRank");
   
    // --------------- the inputs ---------------------

    // page rank input
    InputFormatVertex pageWithRankInput = JobGraphUtils.createInput(new CustomImprovedDanglingPageRankInputFormat(),
      pageWithRankInputPath, "DanglingPageWithRankInput", jobGraph, degreeOfParallelism);
    TaskConfig pageWithRankInputConfig = new TaskConfig(pageWithRankInput.getConfiguration());
    pageWithRankInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    pageWithRankInputConfig.setOutputComparator(vertexWithRankAndDanglingComparator, 0);
    pageWithRankInputConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
    pageWithRankInputConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));

    // edges as adjacency list
    InputFormatVertex adjacencyListInput = JobGraphUtils.createInput(new CustomImprovedAdjacencyListInputFormat(),
      adjacencyListInputPath, "AdjancencyListInput", jobGraph, degreeOfParallelism);
    TaskConfig adjacencyListInputConfig = new TaskConfig(adjacencyListInput.getConfiguration());
    adjacencyListInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    adjacencyListInputConfig.setOutputSerializer(vertexWithAdjacencyListSerializer);
    adjacencyListInputConfig.setOutputComparator(vertexWithAdjacencyListComparator, 0);

    // --------------- the head ---------------------
    AbstractJobVertex head = JobGraphUtils.createTask(IterationHeadPactTask.class, "IterationHead", jobGraph,
      degreeOfParallelism);
    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.setRelativeMemoryInput(0, (double) minorConsumer / totalMemoryConsumption);
    headConfig.setFilehandlesInput(0, NUM_FILE_HANDLES_PER_SORT);
    headConfig.setSpillingThresholdInput(0, SORT_SPILL_THRESHOLD);
   
    // back channel / iterations
    headConfig.setRelativeBackChannelMemory((double) minorConsumer / totalMemoryConsumption);
   
    // 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, new PageRankStatsAggregator());

    // --------------- the join ---------------------
   
    AbstractJobVertex intermediate = JobGraphUtils.createTask(IterationIntermediatePactTask.class,
      "IterationIntermediate", jobGraph, degreeOfParallelism);
    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.setRelativeMemoryDriver((double) matchMemory / totalMemoryConsumption);
    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) --------------------
   
    AbstractJobVertex tail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationTail", jobGraph,
      degreeOfParallelism);
    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.setRelativeInputMaterializationMemory(0, (double)minorConsumer/totalMemoryConsumption);
    tailConfig.setInputLocalStrategy(1, LocalStrategy.SORT);
    tailConfig.setInputComparator(vertexWithRankComparator, 1);
    tailConfig.setRelativeMemoryInput(1, (double) coGroupSortMemory / totalMemoryConsumption);
    tailConfig.setFilehandlesInput(1, NUM_FILE_HANDLES_PER_SORT);
    tailConfig.setSpillingThresholdInput(1, SORT_SPILL_THRESHOLD);
   
    // output
    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 ---------------------

    OutputFormatVertex output = JobGraphUtils.createFileOutput(jobGraph, "FinalOutput", degreeOfParallelism);
    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 ---------------------

    AbstractJobVertex sync = JobGraphUtils.createSync(jobGraph, degreeOfParallelism);
    TaskConfig syncConfig = new TaskConfig(sync.getConfiguration());
    syncConfig.setNumberOfIterations(numIterations);
    syncConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, new PageRankStatsAggregator());
    syncConfig.setConvergenceCriterion(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, new DiffL1NormConvergenceCriterion());
    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(head, sync, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    SlotSharingGroup sharingGroup = new SlotSharingGroup();
    pageWithRankInput.setSlotSharingGroup(sharingGroup);
    adjacencyListInput.setSlotSharingGroup(sharingGroup);
    head.setSlotSharingGroup(sharingGroup);
    intermediate.setSlotSharingGroup(sharingGroup);
    tail.setSlotSharingGroup(sharingGroup);
    output.setSlotSharingGroup(sharingGroup);
    sync.setSlotSharingGroup(sharingGroup);
View Full Code Here
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
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
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
    JobGraph jobGraph = new JobGraph("CompensatableDanglingPageRank");
   
    // --------------- the inputs ---------------------

    // page rank input
    InputFormatVertex pageWithRankInput = JobGraphUtils.createInput(new ImprovedDanglingPageRankInputFormat(),
      pageWithRankInputPath, "DanglingPageWithRankInput", jobGraph, degreeOfParallelism);
    TaskConfig pageWithRankInputConfig = new TaskConfig(pageWithRankInput.getConfiguration());
    pageWithRankInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    pageWithRankInputConfig.setOutputComparator(fieldZeroComparator, 0);
    pageWithRankInputConfig.setOutputSerializer(recSerializer);
    pageWithRankInputConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));

    // edges as adjacency list
    InputFormatVertex adjacencyListInput = JobGraphUtils.createInput(new ImprovedAdjacencyListInputFormat(),
      adjacencyListInputPath, "AdjancencyListInput", jobGraph, degreeOfParallelism);
    TaskConfig adjacencyListInputConfig = new TaskConfig(adjacencyListInput.getConfiguration());
    adjacencyListInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    adjacencyListInputConfig.setOutputSerializer(recSerializer);
    adjacencyListInputConfig.setOutputComparator(fieldZeroComparator, 0);

    // --------------- the head ---------------------
    AbstractJobVertex head = JobGraphUtils.createTask(IterationHeadPactTask.class, "IterationHead", jobGraph, degreeOfParallelism);
    TaskConfig headConfig = new TaskConfig(head.getConfiguration());
    headConfig.setIterationId(ITERATION_ID);
   
    // initial input / partial solution
    headConfig.addInputToGroup(0);
    headConfig.setIterationHeadPartialSolutionOrWorksetInputIndex(0);
    headConfig.setInputSerializer(recSerializer, 0);
    headConfig.setInputComparator(fieldZeroComparator, 0);
    headConfig.setInputLocalStrategy(0, LocalStrategy.SORT);
    headConfig.setRelativeMemoryInput(0, (double)minorConsumer/totalMemoryConsumption);
    headConfig.setFilehandlesInput(0, NUM_FILE_HANDLES_PER_SORT);
    headConfig.setSpillingThresholdInput(0, SORT_SPILL_THRESHOLD);
   
    // back channel / iterations
    headConfig.setRelativeBackChannelMemory((double)minorConsumer/totalMemoryConsumption);
   
    // output into iteration
    headConfig.setOutputSerializer(recSerializer);
    headConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
    headConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
   
    // final output
    TaskConfig headFinalOutConfig = new TaskConfig(new Configuration());
    headFinalOutConfig.setOutputSerializer(recSerializer);
    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<CompensatingMap>(CompensatingMap.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(CompensatableDotProductCoGroup.AGGREGATOR_NAME, new PageRankStatsAggregator());

    // --------------- the join ---------------------
   
    AbstractJobVertex intermediate = JobGraphUtils.createTask(IterationIntermediatePactTask.class, "IterationIntermediate", jobGraph, degreeOfParallelism);
    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.setRelativeMemoryDriver((double)matchMemory/totalMemoryConsumption);
    intermediateConfig.addInputToGroup(0);
    intermediateConfig.addInputToGroup(1);
    intermediateConfig.setInputSerializer(recSerializer, 0);
    intermediateConfig.setInputSerializer(recSerializer, 1);
    intermediateConfig.setDriverComparator(fieldZeroComparator, 0);
    intermediateConfig.setDriverComparator(fieldZeroComparator, 1);
    intermediateConfig.setDriverPairComparator(pairComparatorFactory);
   
    intermediateConfig.setOutputSerializer(recSerializer);
    intermediateConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    intermediateConfig.setOutputComparator(fieldZeroComparator, 0);
   
    intermediateConfig.setStubWrapper(new UserCodeClassWrapper<CompensatableDotProductMatch>(CompensatableDotProductMatch.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) --------------------
   
    AbstractJobVertex tail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationTail", jobGraph,
      degreeOfParallelism);
    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(recSerializer, 0);
    tailConfig.setInputSerializer(recSerializer, 1);
    tailConfig.setDriverComparator(fieldZeroComparator, 0);
    tailConfig.setDriverComparator(fieldZeroComparator, 1);
    tailConfig.setDriverPairComparator(pairComparatorFactory);
    tailConfig.setInputAsynchronouslyMaterialized(0, true);
    tailConfig.setRelativeInputMaterializationMemory(0, (double)minorConsumer/totalMemoryConsumption);
    tailConfig.setInputLocalStrategy(1, LocalStrategy.SORT);
    tailConfig.setInputComparator(fieldZeroComparator, 1);
    tailConfig.setRelativeMemoryInput(1, (double)coGroupSortMemory/totalMemoryConsumption);
    tailConfig.setFilehandlesInput(1, NUM_FILE_HANDLES_PER_SORT);
    tailConfig.setSpillingThresholdInput(1, SORT_SPILL_THRESHOLD);
   
    // output
    tailConfig.setOutputSerializer(recSerializer);
   
    // the stub
    tailConfig.setStubWrapper(new UserCodeClassWrapper<CompensatableDotProductCoGroup>(CompensatableDotProductCoGroup.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 ---------------------

    OutputFormatVertex output = JobGraphUtils.createFileOutput(jobGraph, "FinalOutput", degreeOfParallelism);
    TaskConfig outputConfig = new TaskConfig(output.getConfiguration());
    outputConfig.addInputToGroup(0);
    outputConfig.setInputSerializer(recSerializer, 0);
    outputConfig.setStubWrapper(new UserCodeClassWrapper<PageWithRankOutFormat>(PageWithRankOutFormat.class));
    outputConfig.setStubParameter(FileOutputFormat.FILE_PARAMETER_KEY, outputPath);
   
    // --------------- the auxiliaries ---------------------

    AbstractJobVertex sync = JobGraphUtils.createSync(jobGraph, degreeOfParallelism);
    TaskConfig syncConfig = new TaskConfig(sync.getConfiguration());
    syncConfig.setNumberOfIterations(numIterations);
    syncConfig.addIterationAggregator(CompensatableDotProductCoGroup.AGGREGATOR_NAME, new PageRankStatsAggregator());
    syncConfig.setConvergenceCriterion(CompensatableDotProductCoGroup.AGGREGATOR_NAME, new DiffL1NormConvergenceCriterion());
    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(head, sync, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    SlotSharingGroup sharingGroup = new SlotSharingGroup();
    pageWithRankInput.setSlotSharingGroup(sharingGroup);
    adjacencyListInput.setSlotSharingGroup(sharingGroup);
    head.setSlotSharingGroup(sharingGroup);
    intermediate.setSlotSharingGroup(sharingGroup);
    tail.setSlotSharingGroup(sharingGroup);
    output.setSlotSharingGroup(sharingGroup);
    sync.setSlotSharingGroup(sharingGroup);
View Full Code Here
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
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
322
323
    JobGraph jobGraph = new JobGraph("CompensatableDanglingPageRank");
   
    // --------------- the inputs ---------------------

    // page rank input
    InputFormatVertex pageWithRankInput = JobGraphUtils.createInput(new CustomImprovedDanglingPageRankInputFormat(),
      pageWithRankInputPath, "DanglingPageWithRankInput", jobGraph, degreeOfParallelism);
    TaskConfig pageWithRankInputConfig = new TaskConfig(pageWithRankInput.getConfiguration());
    pageWithRankInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    pageWithRankInputConfig.setOutputComparator(vertexWithRankAndDanglingComparator, 0);
    pageWithRankInputConfig.setOutputSerializer(vertexWithRankAndDanglingSerializer);
    pageWithRankInputConfig.setStubParameter("pageRank.numVertices", String.valueOf(numVertices));

    // edges as adjacency list
    InputFormatVertex adjacencyListInput = JobGraphUtils.createInput(new CustomImprovedAdjacencyListInputFormat(),
      adjacencyListInputPath, "AdjancencyListInput", jobGraph, degreeOfParallelism);
    TaskConfig adjacencyListInputConfig = new TaskConfig(adjacencyListInput.getConfiguration());
    adjacencyListInputConfig.addOutputShipStrategy(ShipStrategyType.PARTITION_HASH);
    adjacencyListInputConfig.setOutputSerializer(vertexWithAdjacencyListSerializer);
    adjacencyListInputConfig.setOutputComparator(vertexWithAdjacencyListComparator, 0);

    // --------------- the head ---------------------
    AbstractJobVertex head = JobGraphUtils.createTask(IterationHeadPactTask.class, "IterationHead", jobGraph,
      degreeOfParallelism);
    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.setRelativeMemoryInput(0, (double)minorConsumer/totalMemoryConsumption);
    headConfig.setFilehandlesInput(0, NUM_FILE_HANDLES_PER_SORT);
    headConfig.setSpillingThresholdInput(0, SORT_SPILL_THRESHOLD);
   
    // back channel / iterations
    headConfig.setRelativeBackChannelMemory((double)minorConsumer/totalMemoryConsumption);
   
    // 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, new PageRankStatsAggregator());

    // --------------- the join ---------------------
   
    AbstractJobVertex intermediate = JobGraphUtils.createTask(IterationIntermediatePactTask.class,
      "IterationIntermediate", jobGraph, degreeOfParallelism);
    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.setRelativeMemoryDriver((double)matchMemory/totalMemoryConsumption);
    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.setDriverComparator(vertexWithRankComparator, 1);
    combinerConfig.setRelativeMemoryDriver((double)coGroupSortMemory/totalMemoryConsumption);
    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) --------------------
   
    AbstractJobVertex tail = JobGraphUtils.createTask(IterationTailPactTask.class, "IterationTail", jobGraph,
      degreeOfParallelism);
    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.setRelativeInputMaterializationMemory(0, (double)minorConsumer/totalMemoryConsumption);
    tailConfig.setInputLocalStrategy(1, LocalStrategy.SORT);
    tailConfig.setInputComparator(vertexWithRankComparator, 1);
    tailConfig.setRelativeMemoryInput(1, (double)coGroupSortMemory/totalMemoryConsumption);
    tailConfig.setFilehandlesInput(1, NUM_FILE_HANDLES_PER_SORT);
    tailConfig.setSpillingThresholdInput(1, SORT_SPILL_THRESHOLD);
    tailConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, new PageRankStatsAggregator());
   
    // output
    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 ---------------------

    OutputFormatVertex output = JobGraphUtils.createFileOutput(jobGraph, "FinalOutput", degreeOfParallelism);
    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 ---------------------

    AbstractJobVertex sync = JobGraphUtils.createSync(jobGraph, degreeOfParallelism);
    TaskConfig syncConfig = new TaskConfig(sync.getConfiguration());
    syncConfig.setNumberOfIterations(numIterations);
    syncConfig.addIterationAggregator(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, new PageRankStatsAggregator());
    syncConfig.setConvergenceCriterion(CustomCompensatableDotProductCoGroup.AGGREGATOR_NAME, new DiffL1NormConvergenceCriterion());
    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(head, sync, ChannelType.NETWORK, DistributionPattern.POINTWISE);
   
    SlotSharingGroup sharingGroup = new SlotSharingGroup();
    pageWithRankInput.setSlotSharingGroup(sharingGroup);
    adjacencyListInput.setSlotSharingGroup(sharingGroup);
    head.setSlotSharingGroup(sharingGroup);
    intermediate.setSlotSharingGroup(sharingGroup);
    tail.setSlotSharingGroup(sharingGroup);
    output.setSlotSharingGroup(sharingGroup);
    sync.setSlotSharingGroup(sharingGroup);
View Full Code Here
236
237
238
239
240
241
242
243
244
245
246
247
248
249
  // -------------------------------------------------------------------------------------------------------------

  @SuppressWarnings("unchecked")
  private static InputFormatVertex createPointsInput(JobGraph jobGraph, String pointsPath, int numSubTasks, TypeSerializerFactory<?> serializer) {
    CsvInputFormat pointsInFormat = new CsvInputFormat(' ', LongValue.class, LongValue.class, LongValue.class, LongValue.class);
    InputFormatVertex pointsInput = JobGraphUtils.createInput(pointsInFormat, pointsPath, "Input[Points]", jobGraph, numSubTasks);

    {
      TaskConfig taskConfig = new TaskConfig(pointsInput.getConfiguration());
      taskConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      taskConfig.setOutputSerializer(serializer);
    }

    return pointsInput;
View Full Code Here
250
251
252
253
254
255
256
257
258
259
260
261
262
263
  }

  @SuppressWarnings("unchecked")
  private static InputFormatVertex createModelsInput(JobGraph jobGraph, String pointsPath, int numSubTasks, TypeSerializerFactory<?> serializer) {
    CsvInputFormat modelsInFormat = new CsvInputFormat(' ', LongValue.class, LongValue.class, LongValue.class, LongValue.class);
    InputFormatVertex modelsInput = JobGraphUtils.createInput(modelsInFormat, pointsPath, "Input[Models]", jobGraph, numSubTasks);

    {
      TaskConfig taskConfig = new TaskConfig(modelsInput.getConfiguration());
      taskConfig.addOutputShipStrategy(ShipStrategyType.BROADCAST);
      taskConfig.setOutputSerializer(serializer);
    }

    return modelsInput;
View Full Code Here
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
    final TypeSerializerFactory<?> serializer = RecordSerializerFactory.get();

    JobGraph jobGraph = new JobGraph("Distance Builder");

    // -- vertices ---------------------------------------------------------------------------------------------
    InputFormatVertex points = createPointsInput(jobGraph, pointsPath, numSubTasks, serializer);
    InputFormatVertex models = createModelsInput(jobGraph, centersPath, numSubTasks, serializer);
    AbstractJobVertex mapper = createMapper(jobGraph, numSubTasks, serializer);
    OutputFormatVertex output = createOutput(jobGraph, resultPath, numSubTasks, serializer);

    // -- edges ------------------------------------------------------------------------------------------------
    JobGraphUtils.connect(points, mapper, ChannelType.NETWORK, DistributionPattern.POINTWISE);
    JobGraphUtils.connect(models, mapper, ChannelType.NETWORK, DistributionPattern.BIPARTITE);
    JobGraphUtils.connect(mapper, output, ChannelType.NETWORK, DistributionPattern.POINTWISE);

    // -- instance sharing -------------------------------------------------------------------------------------
   
    SlotSharingGroup sharing = new SlotSharingGroup();
   
    points.setSlotSharingGroup(sharing);
    models.setSlotSharingGroup(sharing);
    mapper.setSlotSharingGroup(sharing);
    output.setSlotSharingGroup(sharing);

    return jobGraph;
  }
View Full Code Here
96
97
98
99
100
101
102
103
104
105
106
107
108
  // -------------------------------------------------------------------------------------------------------------

  private static InputFormatVertex createPointsInput(JobGraph jobGraph, String pointsPath, int numSubTasks, TypeSerializerFactory<?> serializer) {
    @SuppressWarnings("unchecked")
    CsvInputFormat pointsInFormat = new CsvInputFormat('|', IntValue.class, DoubleValue.class, DoubleValue.class, DoubleValue.class);
    InputFormatVertex pointsInput = JobGraphUtils.createInput(pointsInFormat, pointsPath, "[Points]", jobGraph, numSubTasks);
    {
      TaskConfig taskConfig = new TaskConfig(pointsInput.getConfiguration());
      taskConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      taskConfig.setOutputSerializer(serializer);
     
      TaskConfig chainedMapper = new TaskConfig(new Configuration());
      chainedMapper.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
View Full Code Here
807
808
809
810
811
812
813
814
815
816
817
818
819
820
    assignDriverResources(node, config);
    return vertex;
  }

  private InputFormatVertex createDataSourceVertex(SourcePlanNode node) throws CompilerException {
    final InputFormatVertex vertex = new InputFormatVertex(node.getNodeName());
    final TaskConfig config = new TaskConfig(vertex.getConfiguration());

    vertex.setInvokableClass(DataSourceTask.class);

    // set user code
    config.setStubWrapper(node.getPactContract().getUserCodeWrapper());
    config.setStubParameters(node.getPactContract().getParameters());
View Full Code Here
117
118
119
120
121
122
123
124
125
126
127
128
129
130
  }

  private static InputFormatVertex createCentersInput(JobGraph jobGraph, String centersPath, int numSubTasks, TypeSerializerFactory<?> serializer) {
    @SuppressWarnings("unchecked")
    CsvInputFormat modelsInFormat = new CsvInputFormat('|', IntValue.class, DoubleValue.class, DoubleValue.class, DoubleValue.class);
    InputFormatVertex modelsInput = JobGraphUtils.createInput(modelsInFormat, centersPath, "[Models]", jobGraph, numSubTasks);

    {
      TaskConfig taskConfig = new TaskConfig(modelsInput.getConfiguration());
      taskConfig.addOutputShipStrategy(ShipStrategyType.FORWARD);
      taskConfig.setOutputSerializer(serializer);

      TaskConfig chainedMapper = new TaskConfig(new Configuration());
      chainedMapper.setDriverStrategy(DriverStrategy.COLLECTOR_MAP);
View Full Code Here
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
    final TypeComparatorFactory<?> int0Comparator = new RecordComparatorFactory(new int[] { 0 }, new Class[] { IntValue.class });

    JobGraph jobGraph = new JobGraph("KMeans Iterative");

    // -- vertices ---------------------------------------------------------------------------------------------
    InputFormatVertex points = createPointsInput(jobGraph, pointsPath, numSubTasks, serializer);
    InputFormatVertex centers = createCentersInput(jobGraph, centersPath, numSubTasks, serializer);
   
    AbstractJobVertex head = createIterationHead(jobGraph, numSubTasks, serializer);
    AbstractJobVertex mapper = createMapper(jobGraph, numSubTasks, serializer, serializer, serializer, int0Comparator);
   
    AbstractJobVertex reducer = createReducer(jobGraph, numSubTasks, serializer, int0Comparator, serializer);
   
    AbstractJobVertex sync = createSync(jobGraph, numIterations, numSubTasks);
   
    OutputFormatVertex 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()).setRelativeInputMaterializationMemory(0,
        MEMORY_FRACTION_PER_CONSUMER);

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

    // -- instance sharing -------------------------------------------------------------------------------------
   
    SlotSharingGroup sharingGroup = new SlotSharingGroup();
   
    points.setSlotSharingGroup(sharingGroup);
    centers.setSlotSharingGroup(sharingGroup);
    head.setSlotSharingGroup(sharingGroup);
    mapper.setSlotSharingGroup(sharingGroup);
    reducer.setSlotSharingGroup(sharingGroup);
    sync.setSlotSharingGroup(sharingGroup);
    output.setSlotSharingGroup(sharingGroup);
View Full Code Here
TOP

Related Classes of org.apache.flink.runtime.jobgraph.InputFormatVertex

  • org.apache.flink.compiler.plantranslate.NepheleJobGraphGenerator
  • org.apache.flink.runtime.operators.util.TaskConfig
  • org.apache.flink.test.broadcastvars.BroadcastVarsNepheleITCase
  • org.apache.flink.test.broadcastvars.KMeansIterativeNepheleITCase
  • org.apache.flink.test.iterative.nephele.ConnectedComponentsNepheleITCase
  • org.apache.flink.test.iterative.nephele.customdanglingpagerank.CustomCompensatableDanglingPageRank
  • org.apache.flink.test.iterative.nephele.customdanglingpagerank.CustomCompensatableDanglingPageRankWithCombiner
  • org.apache.flink.test.iterative.nephele.danglingpagerank.CompensatableDanglingPageRank
  • org.apache.flink.test.iterative.nephele.IterationWithChainingNepheleITCase
  • org.apache.flink.test.iterative.nephele.JobGraphUtils
    • 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.