org.apache.drill.exec.planner.physical.DirectScanPrel Maven / Gradle / Ivy

Go to download
/*
 * Licensed to the Apache Software Foundation (ASF) under one
 * or more contributor license agreements.  See the NOTICE file
 * distributed with this work for additional information
 * regarding copyright ownership.  The ASF licenses this file
 * to you under the Apache License, Version 2.0 (the
 * "License"); you may not use this file except in compliance
 * with the License.  You may obtain a copy of the License at
 *
 * http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */
package org.apache.drill.exec.planner.physical;

import java.io.IOException;
import java.util.Collections;
import java.util.Iterator;
import java.util.List;

import org.apache.calcite.plan.RelOptCluster;
import org.apache.calcite.plan.RelOptCost;
import org.apache.calcite.plan.RelOptPlanner;
import org.apache.calcite.plan.RelTraitSet;
import org.apache.calcite.rel.AbstractRelNode;
import org.apache.calcite.rel.RelNode;
import org.apache.calcite.rel.RelWriter;
import org.apache.calcite.rel.metadata.RelMetadataQuery;
import org.apache.calcite.rel.type.RelDataType;
import org.apache.drill.common.exceptions.DrillRuntimeException;
import org.apache.drill.common.exceptions.ExecutionSetupException;
import org.apache.drill.exec.physical.base.GroupScan;
import org.apache.drill.exec.physical.base.PhysicalOperator;
import org.apache.drill.exec.physical.base.ScanStats;
import org.apache.drill.exec.planner.fragment.DistributionAffinity;
import org.apache.drill.exec.planner.physical.visitor.PrelVisitor;
import org.apache.drill.exec.record.BatchSchema;

import org.apache.drill.exec.planner.cost.DrillCostBase.DrillCostFactory;

public class DirectScanPrel extends AbstractRelNode implements Prel, HasDistributionAffinity {
  static final org.slf4j.Logger logger = org.slf4j.LoggerFactory.getLogger(DirectScanPrel.class);

  private final GroupScan groupScan;
  private final RelDataType rowType;

  DirectScanPrel(RelOptCluster cluster, RelTraitSet traits,
                 GroupScan groupScan, RelDataType rowType) {
    super(cluster, traits);
    this.groupScan = groupScan;
    this.rowType = rowType;
  }

  @Override
  public RelNode copy(RelTraitSet traitSet, List inputs) {
    return new DirectScanPrel(this.getCluster(), traitSet, this.getGroupScan(),
            this.rowType);
  }

  @Override
  protected Object clone() throws CloneNotSupportedException {
    return new DirectScanPrel(this.getCluster(), this.getTraitSet(), getCopy(this.getGroupScan()),
            this.rowType);
  }

  private static GroupScan getCopy(GroupScan scan){
    try {
      return (GroupScan) scan.getNewWithChildren((List) (Object) Collections.emptyList());
    } catch (ExecutionSetupException e) {
      throw new DrillRuntimeException("Unexpected failure while coping node.", e);
    }
  }

  @Override
  public PhysicalOperator getPhysicalOperator(PhysicalPlanCreator creator)
          throws IOException {
    return creator.addMetadata(this, this.getGroupScan());
  }

  public GroupScan getGroupScan() {
    return groupScan;
  }

  @Override
  public boolean needsFinalColumnReordering() {
    return false;
  }

  @Override
  public BatchSchema.SelectionVectorMode[] getSupportedEncodings() {
    return BatchSchema.SelectionVectorMode.DEFAULT;
  }

  @Override
  public DistributionAffinity getDistributionAffinity() {
    return this.getGroupScan().getDistributionAffinity();
  }

  @Override
  public BatchSchema.SelectionVectorMode getEncoding() {
    return BatchSchema.SelectionVectorMode.NONE;
  }

  @Override
  public  T accept(PrelVisitor logicalVisitor, X value) throws E {
    return logicalVisitor.visitPrel(this, value);
  }

  @Override
  public Iterator iterator() {
    return Collections.emptyIterator();
  }

  public static DirectScanPrel create(RelNode old, RelTraitSet traitSets,
      GroupScan scan, RelDataType rowType) {
    return new DirectScanPrel(old.getCluster(), traitSets,
        getCopy(scan), rowType);
  }

  @Override
  public RelDataType deriveRowType() {
    return this.rowType;
  }

  @Override
  public RelWriter explainTerms(RelWriter pw) {
    return super.explainTerms(pw).item("groupscan", this.getGroupScan().getDigest());
  }

  @Override
  public double estimateRowCount(RelMetadataQuery mq) {
    final PlannerSettings settings = PrelUtil.getPlannerSettings(getCluster());

    double rowCount = this.getGroupScan().getScanStats(settings).getRecordCount();
    logger.debug("#{}.estimateRowCount get rowCount {} from  groupscan {}",
            this.getId(), rowCount, System.identityHashCode(this.getGroupScan()));
    return rowCount;
  }

  @Override
  public RelOptCost computeSelfCost(final RelOptPlanner planner, RelMetadataQuery mq) {
    final PlannerSettings settings = PrelUtil.getPlannerSettings(planner);
    final ScanStats stats = this.getGroupScan().getScanStats(settings);
    final int columnCount = this.getRowType().getFieldCount();

    if(PrelUtil.getSettings(getCluster()).useDefaultCosting()) {
      return planner.getCostFactory().makeCost(stats.getRecordCount() * columnCount, stats.getCpuCost(), stats.getDiskCost());
    }

    double rowCount = stats.getRecordCount();

    // As DRILL-4083 points out, when columnCount == 0, cpuCost becomes zero,
    // which makes the costs of HiveScan and HiveDrillNativeParquetScan the same
    // For now, assume cpu cost is proportional to row count.
    double cpuCost = rowCount * Math.max(columnCount, 1);

    // If a positive value for CPU cost is given multiply the default CPU cost by given CPU cost.
    if (stats.getCpuCost() > 0) {
      cpuCost *= stats.getCpuCost();
    }

    // Even though scan is reading from disk, in the currently generated plans all plans will
    // need to read the same amount of data, so keeping the disk io cost 0 is ok for now.
    // In the future we might consider alternative scans that go against projections or
    // different compression schemes etc that affect the amount of data read. Such alternatives
    // would affect both cpu and io cost.
    //double ioCost = 0;
    double ioCost = stats.getDiskCost();
    DrillCostFactory costFactory = (DrillCostFactory)planner.getCostFactory();
    return costFactory.makeCost(rowCount, cpuCost, ioCost, 0);
  }

}