Package org.apache.hadoop.hive.ql.plan

Examples of org.apache.hadoop.hive.ql.plan.MapredWork

152
153
154
155
156
157
158
159
160
161
162
      String hiveConfArgs = generateCmdLine(conf);

      // write out the plan to a local file
      Path planPath = new Path(ctx.getLocalTmpFileURI(), "plan.xml");
      OutputStream out = FileSystem.getLocal(conf).create(planPath);
      MapredWork plan = getWork();
      LOG.info("Generating plan file " + planPath.toString());
      Utilities.serializeMapRedWork(plan, out);

      String isSilent = "true".equalsIgnoreCase(System
          .getProperty("test.silent")) ? "-nolog" : "";
View Full Code Here
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
      throws SemanticException {
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = opProcCtx.getMapCurrCtx();
    GenMapRedCtx mapredCtx = mapCurrCtx.get(op.getParentOperators().get(0));
    Task<? extends Serializable> currTask = mapredCtx.getCurrTask();
    MapredWork plan = (MapredWork) currTask.getWork();
    HashMap<Operator<? extends Serializable>, Task<? extends Serializable>> opTaskMap =
      opProcCtx.getOpTaskMap();
    Operator<? extends Serializable> currTopOp = opProcCtx.getCurrTopOp();

    opTaskMap.put(reducer, currTask);
    plan.setReducer(reducer);
    ReduceSinkDesc desc = op.getConf();

    plan.setNumReduceTasks(desc.getNumReducers());

    List<Task<? extends Serializable>> rootTasks = opProcCtx.getRootTasks();

    if (!rootTasks.contains(currTask)) {
        rootTasks.add(currTask);
    }
    if (reducer.getClass() == JoinOperator.class) {
      plan.setNeedsTagging(true);
    }

    assert currTopOp != null;
    List<Operator<? extends Serializable>> seenOps = opProcCtx.getSeenOps();
    String currAliasId = opProcCtx.getCurrAliasId();
View Full Code Here
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
    assert (((pos == -1) && (readInputMapJoin)) || (pos != -1));
    int parentPos = (pos == -1) ? 0 : pos;
    GenMapRedCtx mapredCtx = mapCurrCtx.get(op.getParentOperators().get(
        parentPos));
    Task<? extends Serializable> currTask = mapredCtx.getCurrTask();
    MapredWork plan = (MapredWork) currTask.getWork();
    HashMap<Operator<? extends Serializable>, Task<? extends Serializable>>  opTaskMap =
      opProcCtx.getOpTaskMap();
    Operator<? extends Serializable> currTopOp = opProcCtx.getCurrTopOp();

    // The mapjoin has already been encountered. Some context must be stored
    // about that
    if (readInputMapJoin) {
      AbstractMapJoinOperator<? extends MapJoinDesc> currMapJoinOp = opProcCtx.getCurrMapJoinOp();
      assert currMapJoinOp != null;
      boolean local = ((pos == -1) || (pos == (currMapJoinOp.getConf()).getPosBigTable())) ?
          false : true;

      if (setReducer) {
        Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
        plan.setReducer(reducer);
        opTaskMap.put(reducer, currTask);
        if (reducer.getClass() == JoinOperator.class) {
          plan.setNeedsTagging(true);
        }
        ReduceSinkDesc desc = (ReduceSinkDesc) op.getConf();
        plan.setNumReduceTasks(desc.getNumReducers());
      } else {
        opTaskMap.put(op, currTask);
      }

      if (!readInputUnion) {
View Full Code Here
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
      GenMRProcContext opProcCtx) throws SemanticException {
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = opProcCtx.getMapCurrCtx();
    GenMapRedCtx mapredCtx = mapCurrCtx.get(op.getParentOperators().get(0));
    Task<? extends Serializable> currTask = mapredCtx.getCurrTask();
    MapredWork plan = (MapredWork) currTask.getWork();
    HashMap<Operator<? extends Serializable>, Task<? extends Serializable>> opTaskMap =
      opProcCtx.getOpTaskMap();

    opTaskMap.put(reducer, currTask);
    plan.setReducer(reducer);
    ReduceSinkDesc desc = op.getConf();

    plan.setNumReduceTasks(desc.getNumReducers());

    if (reducer.getClass() == JoinOperator.class) {
      plan.setNeedsTagging(true);
    }

    initUnionPlan(opProcCtx, currTask, false);
  }
View Full Code Here
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
   * It is a idempotent function to add various intermediate files as the source
   * for the union. The plan has already been created.
   */
  public static void initUnionPlan(GenMRProcContext opProcCtx,
      Task<? extends Serializable> currTask, boolean local) {
    MapredWork plan = (MapredWork) currTask.getWork();
    UnionOperator currUnionOp = opProcCtx.getCurrUnionOp();
    assert currUnionOp != null;
    GenMRUnionCtx uCtx = opProcCtx.getUnionTask(currUnionOp);
    assert uCtx != null;

    List<String> taskTmpDirLst = uCtx.getTaskTmpDir();
    List<TableDesc> tt_descLst = uCtx.getTTDesc();
    assert !taskTmpDirLst.isEmpty() && !tt_descLst.isEmpty();
    assert taskTmpDirLst.size() == tt_descLst.size();
    int size = taskTmpDirLst.size();
    assert local == false;

    for (int pos = 0; pos < size; pos++) {
      String taskTmpDir = taskTmpDirLst.get(pos);
      TableDesc tt_desc = tt_descLst.get(pos);
      if (plan.getPathToAliases().get(taskTmpDir) == null) {
        plan.getPathToAliases().put(taskTmpDir, new ArrayList<String>());
        plan.getPathToAliases().get(taskTmpDir).add(taskTmpDir);
        plan.getPathToPartitionInfo().put(taskTmpDir,
            new PartitionDesc(tt_desc, null));
        plan.getAliasToWork().put(taskTmpDir, currUnionOp);
      }
    }
  }
View Full Code Here
371
372
373
374
375
376
377
378
379
380
381
      Task<? extends Serializable> oldTask, Task<? extends Serializable> task,
      GenMRProcContext opProcCtx, int pos, boolean split,
      boolean readMapJoinData, boolean readUnionData, boolean createLocalWork)
      throws SemanticException {
    Task<? extends Serializable> currTask = task;
    MapredWork plan = (MapredWork) currTask.getWork();
    Operator<? extends Serializable> currTopOp = opProcCtx.getCurrTopOp();
    List<Task<? extends Serializable>> parTasks = null;

    // terminate the old task and make current task dependent on it
    if (split) {
View Full Code Here
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
   */
  public static void splitPlan(ReduceSinkOperator op, GenMRProcContext opProcCtx)
  throws SemanticException {
    // Generate a new task
    ParseContext parseCtx = opProcCtx.getParseCtx();
    MapredWork cplan = getMapRedWork(parseCtx.getConf());
    Task<? extends Serializable> redTask = TaskFactory.get(cplan, parseCtx
        .getConf());
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);

    // Add the reducer
    cplan.setReducer(reducer);
    ReduceSinkDesc desc = op.getConf();

    cplan.setNumReduceTasks(new Integer(desc.getNumReducers()));

    HashMap<Operator<? extends Serializable>, Task<? extends Serializable>> opTaskMap =
      opProcCtx.getOpTaskMap();
    opTaskMap.put(reducer, redTask);
    Task<? extends Serializable> currTask = opProcCtx.getCurrTask();
View Full Code Here
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
   * create a new plan and return.
   *
   * @return the new plan
   */
  public static MapredWork getMapRedWork(HiveConf conf) {
    MapredWork work = new MapredWork();
    work.setPathToAliases(new LinkedHashMap<String, ArrayList<String>>());
    work.setPathToPartitionInfo(new LinkedHashMap<String, PartitionDesc>());
    work.setAliasToWork(new LinkedHashMap<String, Operator<? extends Serializable>>());
    work.setTagToValueDesc(new ArrayList<TableDesc>());
    work.setReducer(null);
    work.setHadoopSupportsSplittable(
        conf.getBoolVar(HiveConf.ConfVars.HIVE_COMBINE_INPUT_FORMAT_SUPPORTS_SPLITTABLE));
    return work;
  }
View Full Code Here
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = opProcCtx.getMapCurrCtx();
    mapCurrCtx.put(ts_op, new GenMapRedCtx(childTask, null, null));

    String streamDesc = taskTmpDir;
    MapredWork cplan = (MapredWork) childTask.getWork();

    if (setReducer) {
      Operator<? extends Serializable> reducer = op.getChildOperators().get(0);

      if (reducer.getClass() == JoinOperator.class) {
        String origStreamDesc;
        streamDesc = "$INTNAME";
        origStreamDesc = streamDesc;
        int pos = 0;
        while (cplan.getAliasToWork().get(streamDesc) != null) {
          streamDesc = origStreamDesc.concat(String.valueOf(++pos));
        }
      }

      // TODO: Allocate work to remove the temporary files and make that
      // dependent on the redTask
      if (reducer.getClass() == JoinOperator.class) {
        cplan.setNeedsTagging(true);
      }
    }

    // Add the path to alias mapping
    setTaskPlan(taskTmpDir, streamDesc, ts_op, cplan, local, tt_desc);
View Full Code Here
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
    GenMRUnionCtx uCtxTask = ctx.getUnionTask(union);
    Task<? extends Serializable> uTask = null;

    union.getParentOperators().get(pos);
    MapredWork uPlan = null;

    // union is encountered for the first time
    if (uCtxTask == null) {
      uCtxTask = new GenMRUnionCtx();
      uPlan = GenMapRedUtils.getMapRedWork(parseCtx.getConf());
      uTask = TaskFactory.get(uPlan, parseCtx.getConf());
      uCtxTask.setUTask(uTask);
      ctx.setUnionTask(union, uCtxTask);
    } else {
      uTask = uCtxTask.getUTask();
      uPlan = (MapredWork) uTask.getWork();
    }

    // If there is a mapjoin at position 'pos'
    if (uPrsCtx.getMapJoinSubq(pos)) {
      GenMRMapJoinCtx mjCtx = ctx.getMapJoinCtx(ctx.getCurrMapJoinOp());
      String taskTmpDir = mjCtx.getTaskTmpDir();
      if (uPlan.getPathToAliases().get(taskTmpDir) == null) {
        uPlan.getPathToAliases().put(taskTmpDir, new ArrayList<String>());
        uPlan.getPathToAliases().get(taskTmpDir).add(taskTmpDir);
        uPlan.getPathToPartitionInfo().put(taskTmpDir,
            new PartitionDesc(mjCtx.getTTDesc(), null));
        uPlan.getAliasToWork().put(taskTmpDir, mjCtx.getRootMapJoinOp());
      }

      for (Task t : currTask.getParentTasks()) {
        t.addDependentTask(uTask);
      }
View Full Code Here
TOP

Related Classes of org.apache.hadoop.hive.ql.plan.MapredWork

  • com.sap.hadoop.windowing.runtime2.mr.MRExecutor
  • com.twitter.ambrose.hive.HiveDAGTransformer
  • org.apache.hadoop.hive.ql.exec.ExecDriver
  • org.apache.hadoop.hive.ql.exec.ExecMapper
  • org.apache.hadoop.hive.ql.exec.ExecReducer
  • org.apache.hadoop.hive.ql.exec.MapRedTask
  • org.apache.hadoop.hive.ql.exec.MoveTask
  • org.apache.hadoop.hive.ql.exec.mr.ExecDriver
  • org.apache.hadoop.hive.ql.exec.mr.MapRedTask
  • org.apache.hadoop.hive.ql.exec.TestOperators
    • 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.