Many resources are needed to download a project. Please understand that we have to compensate our server costs. Thank you in advance. Project price only 1 $
You can buy this project and download/modify it how often you want.
/**
* 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.hadoop.hive.ql.plan;
import java.util.ArrayList;
import java.util.EnumSet;
import java.util.List;
import org.apache.hadoop.hive.ql.io.AcidUtils;
import org.apache.hadoop.hive.ql.plan.Explain.Level;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
/**
* ReduceSinkDesc.
*
*/
@Explain(displayName = "Reduce Output Operator", explainLevels = { Level.USER, Level.DEFAULT, Level.EXTENDED })
public class ReduceSinkDesc extends AbstractOperatorDesc {
private static final long serialVersionUID = 1L;
/**
* Key columns are passed to reducer in the "key".
*/
private java.util.ArrayList keyCols;
private java.util.ArrayList outputKeyColumnNames;
private List> distinctColumnIndices;
/**
* Value columns are passed to reducer in the "value".
*/
private java.util.ArrayList valueCols;
private java.util.ArrayList outputValueColumnNames;
/**
* Describe how to serialize the key.
*/
private TableDesc keySerializeInfo;
/**
* Describe how to serialize the value.
*/
private TableDesc valueSerializeInfo;
/**
* The tag for this reducesink descriptor.
*/
private int tag;
/**
* Number of distribution keys.
*/
private int numDistributionKeys;
/**
* Used in tez. Holds the name of the output
* that this reduce sink is writing to.
*/
private String outputName;
/**
* Holds the name of the output operators
* that this reduce sink is outputing to.
*/
private List outputOperators;
/**
* The partition columns (CLUSTER BY or DISTRIBUTE BY in Hive language).
* Partition columns decide the reducer that the current row goes to.
* Partition columns are not passed to reducer.
*/
private java.util.ArrayList partitionCols;
private int numReducers;
/**
* Bucket information
*/
private int numBuckets;
private List bucketCols;
private int topN = -1;
private float topNMemoryUsage = -1;
private boolean mapGroupBy; // for group-by, values with same key on top-K should be forwarded
//flag used to control how TopN handled for PTF/Windowing partitions.
private boolean isPTFReduceSink = false;
private boolean skipTag; // Skip writing tags when feeding into mapjoin hashtable
public static enum ReducerTraits {
UNSET(0), // unset
FIXED(1), // distribution of keys is fixed
AUTOPARALLEL(2), // can change reducer count (ORDER BY can concat adjacent buckets)
UNIFORM(3); // can redistribute into buckets uniformly (GROUP BY can)
private final int trait;
private ReducerTraits(int trait) {
this.trait = trait;
}
};
// Is reducer auto-parallelism unset (FIXED, UNIFORM, PARALLEL)
private EnumSet reduceTraits = EnumSet.of(ReducerTraits.UNSET);
// Write type, since this needs to calculate buckets differently for updates and deletes
private AcidUtils.Operation writeType;
// whether this RS is deduplicated
private transient boolean isDeduplicated = false;
// used by spark mode to decide whether global order is needed
private transient boolean hasOrderBy = false;
private static transient Logger LOG = LoggerFactory.getLogger(ReduceSinkDesc.class);
// Extra parameters only for vectorization.
private VectorReduceSinkDesc vectorDesc;
public ReduceSinkDesc() {
}
public ReduceSinkDesc(ArrayList keyCols,
int numDistributionKeys,
ArrayList valueCols,
ArrayList outputKeyColumnNames,
List> distinctColumnIndices,
ArrayList outputValueColumnNames, int tag,
ArrayList partitionCols, int numReducers,
final TableDesc keySerializeInfo, final TableDesc valueSerializeInfo,
AcidUtils.Operation writeType) {
this.keyCols = keyCols;
this.numDistributionKeys = numDistributionKeys;
this.valueCols = valueCols;
this.outputKeyColumnNames = outputKeyColumnNames;
this.outputValueColumnNames = outputValueColumnNames;
this.tag = tag;
this.numReducers = numReducers;
this.partitionCols = partitionCols;
this.keySerializeInfo = keySerializeInfo;
this.valueSerializeInfo = valueSerializeInfo;
this.distinctColumnIndices = distinctColumnIndices;
this.setNumBuckets(-1);
this.setBucketCols(null);
this.writeType = writeType;
this.vectorDesc = null;
}
@Override
public Object clone() {
ReduceSinkDesc desc = new ReduceSinkDesc();
desc.setKeyCols((ArrayList) getKeyCols().clone());
desc.setValueCols((ArrayList) getValueCols().clone());
desc.setOutputKeyColumnNames((ArrayList) getOutputKeyColumnNames().clone());
List> distinctColumnIndicesClone = new ArrayList>();
for (List distinctColumnIndex : getDistinctColumnIndices()) {
List tmp = new ArrayList();
tmp.addAll(distinctColumnIndex);
distinctColumnIndicesClone.add(tmp);
}
desc.setDistinctColumnIndices(distinctColumnIndicesClone);
desc.setOutputValueColumnNames((ArrayList) getOutputValueColumnNames().clone());
desc.setNumDistributionKeys(getNumDistributionKeys());
desc.setTag(getTag());
desc.setNumReducers(getNumReducers());
desc.setPartitionCols((ArrayList) getPartitionCols().clone());
desc.setKeySerializeInfo((TableDesc) getKeySerializeInfo().clone());
desc.setValueSerializeInfo((TableDesc) getValueSerializeInfo().clone());
desc.setNumBuckets(numBuckets);
desc.setBucketCols(bucketCols);
desc.setStatistics(this.getStatistics());
desc.setSkipTag(skipTag);
desc.reduceTraits = reduceTraits.clone();
desc.setDeduplicated(isDeduplicated);
desc.setHasOrderBy(hasOrderBy);
if (vectorDesc != null) {
throw new RuntimeException("Clone with vectorization desc not supported");
}
desc.vectorDesc = null;
return desc;
}
public void setVectorDesc(VectorReduceSinkDesc vectorDesc) {
this.vectorDesc = vectorDesc;
}
public VectorReduceSinkDesc getVectorDesc() {
return vectorDesc;
}
public java.util.ArrayList getOutputKeyColumnNames() {
return outputKeyColumnNames;
}
public void setOutputKeyColumnNames(
java.util.ArrayList outputKeyColumnNames) {
this.outputKeyColumnNames = outputKeyColumnNames;
}
public java.util.ArrayList getOutputValueColumnNames() {
return outputValueColumnNames;
}
public void setOutputValueColumnNames(
java.util.ArrayList outputValueColumnNames) {
this.outputValueColumnNames = outputValueColumnNames;
}
@Explain(displayName = "key expressions")
public String getKeyColString() {
return PlanUtils.getExprListString(keyCols);
}
public java.util.ArrayList getKeyCols() {
return keyCols;
}
public void setKeyCols(final java.util.ArrayList keyCols) {
this.keyCols = keyCols;
}
public int getNumDistributionKeys() {
return this.numDistributionKeys;
}
public void setNumDistributionKeys(int numKeys) {
this.numDistributionKeys = numKeys;
}
@Explain(displayName = "value expressions")
public String getValueColsString() {
return PlanUtils.getExprListString(valueCols);
}
public java.util.ArrayList getValueCols() {
return valueCols;
}
public void setValueCols(final java.util.ArrayList valueCols) {
this.valueCols = valueCols;
}
@Explain(displayName = "Map-reduce partition columns")
public String getParitionColsString() {
return PlanUtils.getExprListString(partitionCols);
}
@Explain(displayName = "PartitionCols", explainLevels = { Level.USER })
public String getUserLevelExplainParitionColsString() {
return PlanUtils.getExprListString(partitionCols, true);
}
public java.util.ArrayList getPartitionCols() {
return partitionCols;
}
public void setPartitionCols(
final java.util.ArrayList partitionCols) {
this.partitionCols = partitionCols;
}
public boolean isPartitioning() {
if (partitionCols != null && !partitionCols.isEmpty()) {
return true;
}
return false;
}
@Explain(displayName = "tag", explainLevels = { Level.EXTENDED })
public int getTag() {
return tag;
}
public void setTag(int tag) {
this.tag = tag;
}
public int getTopN() {
return topN;
}
public void setTopN(int topN) {
this.topN = topN;
}
@Explain(displayName = "TopN", explainLevels = { Level.EXTENDED })
public Integer getTopNExplain() {
return topN > 0 ? topN : null;
}
public float getTopNMemoryUsage() {
return topNMemoryUsage;
}
public void setTopNMemoryUsage(float topNMemoryUsage) {
this.topNMemoryUsage = topNMemoryUsage;
}
@Explain(displayName = "TopN Hash Memory Usage")
public Float getTopNMemoryUsageExplain() {
return topN > 0 && topNMemoryUsage > 0 ? topNMemoryUsage : null;
}
public boolean isMapGroupBy() {
return mapGroupBy;
}
public void setMapGroupBy(boolean mapGroupBy) {
this.mapGroupBy = mapGroupBy;
}
public boolean isPTFReduceSink() {
return isPTFReduceSink;
}
public void setPTFReduceSink(boolean isPTFReduceSink) {
this.isPTFReduceSink = isPTFReduceSink;
}
/**
* Returns the number of reducers for the map-reduce job. -1 means to decide
* the number of reducers at runtime. This enables Hive to estimate the number
* of reducers based on the map-reduce input data size, which is only
* available right before we start the map-reduce job.
*/
public int getNumReducers() {
return numReducers;
}
public void setNumReducers(int numReducers) {
this.numReducers = numReducers;
}
public TableDesc getKeySerializeInfo() {
return keySerializeInfo;
}
public void setKeySerializeInfo(TableDesc keySerializeInfo) {
this.keySerializeInfo = keySerializeInfo;
}
public TableDesc getValueSerializeInfo() {
return valueSerializeInfo;
}
public void setValueSerializeInfo(TableDesc valueSerializeInfo) {
this.valueSerializeInfo = valueSerializeInfo;
}
/**
* Returns the sort order of the key columns.
*
* @return null, which means ascending order for all key columns, or a String
* of the same length as key columns, that consists of only "+"
* (ascending order) and "-" (descending order).
*/
@Explain(displayName = "sort order")
public String getOrder() {
return keySerializeInfo.getProperties().getProperty(
org.apache.hadoop.hive.serde.serdeConstants.SERIALIZATION_SORT_ORDER);
}
public void setOrder(String orderStr) {
keySerializeInfo.getProperties().setProperty(
org.apache.hadoop.hive.serde.serdeConstants.SERIALIZATION_SORT_ORDER,
orderStr);
}
public boolean isOrdering() {
if (this.getOrder() != null && !this.getOrder().isEmpty()) {
return true;
}
return false;
}
/**
* Returns the null order in the key columns.
*
* @return null, which means default for all key columns, or a String
* of the same length as key columns, that consists of only "a"
* (null first) and "z" (null last).
*/
@Explain(displayName = "null sort order", explainLevels = { Level.EXTENDED })
public String getNullOrder() {
return keySerializeInfo.getProperties().getProperty(
org.apache.hadoop.hive.serde.serdeConstants.SERIALIZATION_NULL_SORT_ORDER);
}
public void setNullOrder(String nullOrderStr) {
keySerializeInfo.getProperties().setProperty(
org.apache.hadoop.hive.serde.serdeConstants.SERIALIZATION_NULL_SORT_ORDER,
nullOrderStr);
}
public List> getDistinctColumnIndices() {
return distinctColumnIndices;
}
public void setDistinctColumnIndices(
List> distinctColumnIndices) {
this.distinctColumnIndices = distinctColumnIndices;
}
@Explain(displayName = "outputname", explainLevels = { Level.USER })
public String getOutputName() {
return outputName;
}
public void setOutputName(String outputName) {
this.outputName = outputName;
}
public int getNumBuckets() {
return numBuckets;
}
public void setNumBuckets(int numBuckets) {
this.numBuckets = numBuckets;
}
public List getBucketCols() {
return bucketCols;
}
public void setBucketCols(List bucketCols) {
this.bucketCols = bucketCols;
}
public void setSkipTag(boolean value) {
this.skipTag = value;
}
public boolean getSkipTag() {
return skipTag;
}
@Explain(displayName = "auto parallelism", explainLevels = { Level.EXTENDED })
public final boolean isAutoParallel() {
return (this.reduceTraits.contains(ReducerTraits.AUTOPARALLEL));
}
public final EnumSet getReducerTraits() {
return this.reduceTraits;
}
public final void setReducerTraits(EnumSet traits) {
// we don't allow turning on auto parallel once it has been
// explicitly turned off. That is to avoid scenarios where
// auto parallelism could break assumptions about number of
// reducers or hash function.
boolean wasUnset = this.reduceTraits.remove(ReducerTraits.UNSET);
if (this.reduceTraits.contains(ReducerTraits.FIXED)) {
return;
} else if (traits.contains(ReducerTraits.FIXED)) {
this.reduceTraits.removeAll(EnumSet.of(
ReducerTraits.AUTOPARALLEL,
ReducerTraits.UNIFORM));
this.reduceTraits.addAll(traits);
} else {
this.reduceTraits.addAll(traits);
}
}
public AcidUtils.Operation getWriteType() {
return writeType;
}
public boolean isDeduplicated() {
return isDeduplicated;
}
public void setDeduplicated(boolean isDeduplicated) {
this.isDeduplicated = isDeduplicated;
}
public boolean hasOrderBy() {
return hasOrderBy;
}
public void setHasOrderBy(boolean hasOrderBy) {
this.hasOrderBy = hasOrderBy;
}
public List getOutputOperators() {
return outputOperators;
}
public void setOutputOperators(List outputOperators) {
this.outputOperators = outputOperators;
}
}
