Package org.apache.hadoop.hive.ql.exec

Examples of org.apache.hadoop.hive.ql.exec.ReduceSinkOperator

548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
  public static void setKeyAndValueDesc(mapredWork plan, Operator<? extends Serializable> topOp) {
    if (topOp == null)
      return;
   
    if (topOp instanceof ReduceSinkOperator) {
      ReduceSinkOperator rs = (ReduceSinkOperator)topOp;
      plan.setKeyDesc(rs.getConf().getKeySerializeInfo());
      int tag = Math.max(0, rs.getConf().getTag());
      List<tableDesc> tagToSchema = plan.getTagToValueDesc();
      while (tag + 1 > tagToSchema.size()) {
        tagToSchema.add(null);
      }
      tagToSchema.set(tag, rs.getConf().getValueSerializeInfo());
    } else {
      List<Operator<? extends Serializable>> children = topOp.getChildOperators();
      if (children != null) {
        for(Operator<? extends Serializable> op: children) {
          setKeyAndValueDesc(plan, op);
View Full Code Here
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
  /**
   * The Node Processor for Column Pruning on Reduce Sink Operators.
   */
  public static class ColumnPrunerReduceSinkProc implements NodeProcessor {
    public Object process(Node nd, Stack<Node> stack, NodeProcessorCtx ctx, Object... nodeOutputs) throws SemanticException {
      ReduceSinkOperator op = (ReduceSinkOperator)nd;
      ColumnPrunerProcCtx cppCtx = (ColumnPrunerProcCtx)ctx;
      HashMap<Operator<? extends Serializable>, OpParseContext> opToParseCtxMap =
          cppCtx.getOpToParseCtxMap();
      RowResolver redSinkRR = opToParseCtxMap.get(op).getRR();
      reduceSinkDesc conf = op.getConf();
      List<Operator<? extends Serializable>> childOperators = op.getChildOperators();
      List<Operator<? extends Serializable>> parentOperators = op.getParentOperators();

      List<String> colLists = new ArrayList<String>();
      ArrayList<exprNodeDesc> keys = conf.getKeyCols();
      for (exprNodeDesc key : keys)
        colLists = Utilities.mergeUniqElems(colLists, key.getCols());

      if ((childOperators.size() == 1) && (childOperators.get(0) instanceof JoinOperator)) {
        assert parentOperators.size() == 1;
        Operator<? extends Serializable> par = parentOperators.get(0);
        JoinOperator childJoin = (JoinOperator)childOperators.get(0);
        RowResolver parRR = opToParseCtxMap.get(par).getRR();
        List<String> childJoinCols = cppCtx.getJoinPrunedColLists().get(childJoin).get((byte)conf.getTag());
        boolean[] flags = new boolean[conf.getValueCols().size()];
        for (int i = 0; i < flags.length; i++)
          flags[i] = false;
        if (childJoinCols != null && childJoinCols.size() > 0) {
          Map<String,exprNodeDesc> exprMap = op.getColumnExprMap();
          for (String childCol : childJoinCols) {
            exprNodeDesc desc = exprMap.get(childCol);
            int index = conf.getValueCols().indexOf(desc);
            flags[index] = true;
            String[] nm = redSinkRR.reverseLookup(childCol);
View Full Code Here
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
   * Reduce Scan encountered
   * @param nd the reduce sink operator encountered
   * @param opProcCtx context
   */
  public Object process(Node nd, Stack<Node> stack, NodeProcessorCtx opProcCtx, Object... nodeOutputs) throws SemanticException {
    ReduceSinkOperator op = (ReduceSinkOperator)nd;
    GenMRProcContext ctx = (GenMRProcContext)opProcCtx;

    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = ctx.getMapCurrCtx();
    GenMapRedCtx mapredCtx = mapCurrCtx.get((Operator<? extends Serializable>)stack.get(stack.size()-2));
    Task<? extends Serializable> currTask    = mapredCtx.getCurrTask();
    mapredWork currPlan = (mapredWork) currTask.getWork();
    Operator<? extends Serializable> currTopOp   = mapredCtx.getCurrTopOp();
    String currAliasId = mapredCtx.getCurrAliasId();
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
    HashMap<Operator<? extends Serializable>, Task<? extends Serializable>> opTaskMap = ctx.getOpTaskMap();
    Task<? extends Serializable> opMapTask = opTaskMap.get(reducer);

    ctx.setCurrTopOp(currTopOp);
    ctx.setCurrAliasId(currAliasId);
View Full Code Here
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
   * Reduce Scan encountered
   * @param nd the reduce sink operator encountered
   * @param opProcCtx context
   */
  public Object process(Node nd, Stack<Node> stack, NodeProcessorCtx opProcCtx, Object... nodeOutputs) throws SemanticException {
    ReduceSinkOperator op = (ReduceSinkOperator)nd;
    GenMRProcContext ctx = (GenMRProcContext)opProcCtx;

    ParseContext parseCtx = ctx.getParseCtx();
    UnionProcContext uCtx = parseCtx.getUCtx();

    // union was map only - no special processing needed
    if (uCtx.isMapOnlySubq())
      return (new GenMRRedSink1()).process(nd, stack, opProcCtx, nodeOutputs);

    // union consisted on a bunch of map-reduce jobs, and it has been split at the union
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = ctx.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 = ctx.getOpTaskMap();
    Task<? extends Serializable> opMapTask = opTaskMap.get(reducer);
   
View Full Code Here
147
148
149
150
151
152
153
154
155
156
157
158
    int keyLength = 0;
   
    //get the join keys from old parent ReduceSink operators
    for (pos = 0; pos < newParentOps.size(); pos++) {
      ReduceSinkOperator oldPar = (ReduceSinkOperator)oldReduceSinkParentOps.get(pos);
      reduceSinkDesc rsconf = oldPar.getConf();
      Byte tag = (byte)rsconf.getTag();
      List<exprNodeDesc> keys = rsconf.getKeyCols();
      keyExprMap.put(tag, keys);
    }
   
View Full Code Here
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
   * Reduce Scan encountered
   * @param nd the reduce sink operator encountered
   * @param opProcCtx context
   */
  public Object process(Node nd, Stack<Node> stack, NodeProcessorCtx opProcCtx, Object... nodeOutputs) throws SemanticException {
    ReduceSinkOperator op = (ReduceSinkOperator)nd;
    GenMRProcContext ctx = (GenMRProcContext)opProcCtx;

    ParseContext parseCtx = ctx.getParseCtx();

    // map-join consisted on a bunch of map-only jobs, and it has been split after the mapjoin
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = ctx.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 = ctx.getOpTaskMap();
    Task<? extends Serializable> opMapTask = opTaskMap.get(reducer);
   
View Full Code Here
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
   * Reduce Scan encountered
   * @param nd the reduce sink operator encountered
   * @param opProcCtx context
   */
  public Object process(Node nd, Stack<Node> stack, NodeProcessorCtx opProcCtx, Object... nodeOutputs) throws SemanticException {
    ReduceSinkOperator op = (ReduceSinkOperator)nd;
    GenMRProcContext ctx = (GenMRProcContext)opProcCtx;

    Map<Operator<? extends Serializable>, GenMapRedCtx> mapCurrCtx = ctx.getMapCurrCtx();
    GenMapRedCtx mapredCtx = mapCurrCtx.get(op.getParentOperators().get(0));
    Task<? extends Serializable> currTask      = mapredCtx.getCurrTask();
    Operator<? extends Serializable> currTopOp = mapredCtx.getCurrTopOp();
    String currAliasId = mapredCtx.getCurrAliasId();
    Operator<? extends Serializable> reducer = op.getChildOperators().get(0);
    Map<Operator<? extends Serializable>, Task<? extends Serializable>> opTaskMap = ctx.getOpTaskMap();
    Task<? extends Serializable> opMapTask = opTaskMap.get(reducer);

    ctx.setCurrTopOp(currTopOp);
    ctx.setCurrAliasId(currAliasId);
View Full Code Here
2035
2036
2037
2038
2039
2040
2041
2042
2043
2044
2045
2046
2047
2048
2049
                                        new ColumnInfo(field,
                                                       type, null, false));
      }
    }

    ReduceSinkOperator rsOp = (ReduceSinkOperator)  putOpInsertMap(
      OperatorFactory.getAndMakeChild(PlanUtils.getReduceSinkDesc(reduceKeys, reduceValues, outputColumnNames, true, -1, numPartitionFields,
                                                                  numReducers),
                                        new RowSchema(reduceSinkOutputRowResolver.getColumnInfos()),
                                        inputOperatorInfo),
        reduceSinkOutputRowResolver
    );
    rsOp.setColumnExprMap(colExprMap);
    return rsOp;
  }
View Full Code Here
2102
2103
2104
2105
2106
2107
2108
2109
2110
2111
2112
2113
2114
2115
2116
2117
      reduceSinkOutputRowResolver2.put("", t.toStringTree(),
                                       new ColumnInfo(Utilities.ReduceField.VALUE.toString() + "." + col,
                                           typeInfo, "", false));
    }

    ReduceSinkOperator rsOp = (ReduceSinkOperator) putOpInsertMap(
      OperatorFactory.getAndMakeChild(PlanUtils.getReduceSinkDesc(reduceKeys, reduceValues, outputColumnNames, true, -1,
                                                                  numPartitionFields, numReducers),
                                        new RowSchema(reduceSinkOutputRowResolver2.getColumnInfos()),
                                        groupByOperatorInfo),
        reduceSinkOutputRowResolver2
    );
   
    rsOp.setColumnExprMap(colExprMap);
    return rsOp;
  }
View Full Code Here
3101
3102
3103
3104
3105
3106
3107
3108
3109
3110
3111
3112
3113
      // Cartesian product is not supported in strict mode
      if (conf.getVar(HiveConf.ConfVars.HIVEMAPREDMODE).equalsIgnoreCase("strict"))
        throw new SemanticException(ErrorMsg.NO_CARTESIAN_PRODUCT.getMsg());
    }

    ReduceSinkOperator rsOp = (ReduceSinkOperator)putOpInsertMap(
      OperatorFactory.getAndMakeChild(
        PlanUtils.getReduceSinkDesc(reduceKeys, reduceValues, outputColumns, false, joinTree.getNextTag(), reduceKeys.size(), numReds),
        new RowSchema(outputRS.getColumnInfos()),
        child), outputRS);
    rsOp.setColumnExprMap(colExprMap);
    return rsOp;
  }
View Full Code Here
TOP

Related Classes of org.apache.hadoop.hive.ql.exec.ReduceSinkOperator

  • org.apache.hadoop.hive.ql.io.HiveKey
  • org.apache.hadoop.hive.ql.optimizer.BucketingSortingReduceSinkOptimizer$BucketSortReduceSinkProcessor
  • org.apache.hadoop.hive.ql.optimizer.ColumnPrunerProcFactory$ColumnPrunerReduceSinkProc
  • org.apache.hadoop.hive.ql.optimizer.ConstantPropagateProcFactory$ConstantPropagateReduceSinkProc
  • org.apache.hadoop.hive.ql.optimizer.ConvertJoinMapJoin
  • org.apache.hadoop.hive.ql.optimizer.correlation.CorrelationOptimizer$CorrelationNodeProc
  • org.apache.hadoop.hive.ql.optimizer.correlation.QueryPlanTreeTransformation
  • org.apache.hadoop.hive.ql.optimizer.correlation.ReduceSinkDeDuplication$AbsctractReducerReducerProc
  • org.apache.hadoop.hive.ql.optimizer.correlation.ReduceSinkDeDuplication$GroupbyReducerProc
  • org.apache.hadoop.hive.ql.optimizer.correlation.ReduceSinkDeDuplication$ReducerReducerProc
    • 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.