[
https://issues.apache.org/jira/browse/HIVE-21196?focusedWorklogId=467224&page=com.atlassian.jira.plugin.system.issuetabpanels:worklog-tabpanel#worklog-467224
]
ASF GitHub Bot logged work on HIVE-21196:
-----------------------------------------
Author: ASF GitHub Bot
Created on: 06/Aug/20 11:31
Start Date: 06/Aug/20 11:31
Worklog Time Spent: 10m
Work Description: zabetak commented on a change in pull request #1325:
URL: https://github.com/apache/hive/pull/1325#discussion_r466347783
##########
File path: ql/src/java/org/apache/hadoop/hive/ql/exec/OperatorUtils.java
##########
@@ -53,6 +53,34 @@
private static final Logger LOG =
LoggerFactory.getLogger(OperatorUtils.class);
+ /**
+ * Return the ancestor of the specified operator at the provided path or
null if the path is invalid.
+ *
+ * The method is equivalent to following code:
+ * <pre>{@code
+ * op.getParentOperators().get(path[0])
+ * .getParentOperators().get(path[1])
+ * ...
+ * .getParentOperators().get(path[n])
+ * }</pre>
+ * with additional checks about the validity of the provided path and the
type of the ancestor.
+ *
+ * @param op the operator for which we
+ * @param clazz the class of the ancestor operator
+ * @param path the path leading to the desired ancestor
+ * @param <T> the type of the ancestor
+ * @return the ancestor of the specified operator at the provided path or
null if the path is invalid.
+ */
+ public static <T> T ancestor(Operator<?> op, Class<T> clazz, int... path) {
+ Operator<?> target = op;
+ for (int i = 0; i < path.length; i++) {
+ if (target.getParentOperators() == null || path[i] >
target.getParentOperators().size())
Review comment:
Done. I also configured IntelliJ to force their usage in single line
statements so hopefully they should never appear.
##########
File path:
ql/src/java/org/apache/hadoop/hive/ql/optimizer/SemiJoinReductionMerge.java
##########
@@ -0,0 +1,399 @@
+/*
+ * 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.optimizer;
+
+import org.apache.hadoop.hive.conf.HiveConf;
+import org.apache.hadoop.hive.ql.exec.ColumnInfo;
+import org.apache.hadoop.hive.ql.exec.FilterOperator;
+import org.apache.hadoop.hive.ql.exec.GroupByOperator;
+import org.apache.hadoop.hive.ql.exec.Operator;
+import org.apache.hadoop.hive.ql.exec.OperatorFactory;
+import org.apache.hadoop.hive.ql.exec.OperatorUtils;
+import org.apache.hadoop.hive.ql.exec.ReduceSinkOperator;
+import org.apache.hadoop.hive.ql.exec.RowSchema;
+import org.apache.hadoop.hive.ql.exec.SelectOperator;
+import org.apache.hadoop.hive.ql.exec.TableScanOperator;
+import org.apache.hadoop.hive.ql.exec.Utilities;
+import org.apache.hadoop.hive.ql.io.AcidUtils;
+import org.apache.hadoop.hive.ql.parse.GenTezUtils;
+import org.apache.hadoop.hive.ql.parse.ParseContext;
+import org.apache.hadoop.hive.ql.parse.RuntimeValuesInfo;
+import org.apache.hadoop.hive.ql.parse.SemanticAnalyzer;
+import org.apache.hadoop.hive.ql.parse.SemanticException;
+import org.apache.hadoop.hive.ql.parse.SemiJoinBranchInfo;
+import org.apache.hadoop.hive.ql.plan.AggregationDesc;
+import org.apache.hadoop.hive.ql.plan.DynamicValue;
+import org.apache.hadoop.hive.ql.plan.ExprNodeColumnDesc;
+import org.apache.hadoop.hive.ql.plan.ExprNodeConstantDesc;
+import org.apache.hadoop.hive.ql.plan.ExprNodeDesc;
+import org.apache.hadoop.hive.ql.plan.ExprNodeDynamicValueDesc;
+import org.apache.hadoop.hive.ql.plan.ExprNodeGenericFuncDesc;
+import org.apache.hadoop.hive.ql.plan.FilterDesc;
+import org.apache.hadoop.hive.ql.plan.GroupByDesc;
+import org.apache.hadoop.hive.ql.plan.PlanUtils;
+import org.apache.hadoop.hive.ql.plan.ReduceSinkDesc;
+import org.apache.hadoop.hive.ql.plan.SelectDesc;
+import org.apache.hadoop.hive.ql.plan.TableDesc;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDAFBloomFilter;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDAFEvaluator;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDAFMax;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDAFMin;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDFBetween;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDFInBloomFilter;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDFMurmurHash;
+import org.apache.hadoop.hive.ql.udf.generic.GenericUDFOPAnd;
+import org.apache.hadoop.hive.ql.util.NullOrdering;
+import org.apache.hadoop.hive.serde2.typeinfo.TypeInfo;
+import org.apache.hadoop.hive.serde2.typeinfo.TypeInfoFactory;
+
+import java.util.ArrayDeque;
+import java.util.ArrayList;
+import java.util.Arrays;
+import java.util.Collections;
+import java.util.Comparator;
+import java.util.Deque;
+import java.util.EnumSet;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+import java.util.SortedMap;
+import java.util.TreeMap;
+
+public class SemiJoinReductionMerge extends Transform {
Review comment:
Done
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
Issue Time Tracking
-------------------
Worklog Id: (was: 467224)
Time Spent: 1h 10m (was: 1h)
> Support semijoin reduction on multiple column join
> --------------------------------------------------
>
> Key: HIVE-21196
> URL: https://issues.apache.org/jira/browse/HIVE-21196
> Project: Hive
> Issue Type: Bug
> Reporter: Deepak Jaiswal
> Assignee: Stamatis Zampetakis
> Priority: Major
> Labels: pull-request-available
> Time Spent: 1h 10m
> Remaining Estimate: 0h
>
> Currently for a query involving join on multiple columns creates separate
> semi join edges for each key which in turn create a bloom filter for each of
> them, like below,
> EXPLAIN select count(*) from srcpart_date_n7 join srcpart_small_n3 on
> (srcpart_date_n7.key = srcpart_small_n3.key1 and srcpart_date_n7.value =
> srcpart_small_n3.value1)
> {code:java}
> Map 1 <- Reducer 5 (BROADCAST_EDGE)
> Reducer 2 <- Map 1 (SIMPLE_EDGE), Map 4 (SIMPLE_EDGE)
> Reducer 3 <- Reducer 2 (CUSTOM_SIMPLE_EDGE)
> Reducer 5 <- Map 4 (CUSTOM_SIMPLE_EDGE)
> #### A masked pattern was here ####
> Vertices:
> Map 1
> Map Operator Tree:
> TableScan
> alias: srcpart_date_n7
> filterExpr: (key is not null and value is not null and (key
> BETWEEN DynamicValue(RS_7_srcpart_small_n3_key1_min) AND
> DynamicValue(RS_7_srcpart_small_n3_key1_max) and in_bloom_filter(key,
> DynamicValue(RS_7_srcpart_small_n3_key1_bloom_filter)))) (type: boolean)
> Statistics: Num rows: 2000 Data size: 356000 Basic stats:
> COMPLETE Column stats: COMPLETE
> Filter Operator
> predicate: ((key BETWEEN
> DynamicValue(RS_7_srcpart_small_n3_key1_min) AND
> DynamicValue(RS_7_srcpart_small_n3_key1_max) and in_bloom_filter(key,
> DynamicValue(RS_7_srcpart_small_n3_key1_bloom_filter))) and key is not null
> and value is not null) (type: boolean)
> Statistics: Num rows: 2000 Data size: 356000 Basic stats:
> COMPLETE Column stats: COMPLETE
> Select Operator
> expressions: key (type: string), value (type: string)
> outputColumnNames: _col0, _col1
> Statistics: Num rows: 2000 Data size: 356000 Basic
> stats: COMPLETE Column stats: COMPLETE
> Reduce Output Operator
> key expressions: _col0 (type: string), _col1 (type:
> string)
> sort order: ++
> Map-reduce partition columns: _col0 (type: string),
> _col1 (type: string)
> Statistics: Num rows: 2000 Data size: 356000 Basic
> stats: COMPLETE Column stats: COMPLETE
> Execution mode: vectorized, llap
> LLAP IO: all inputs
> Map 4
> Map Operator Tree:
> TableScan
> alias: srcpart_small_n3
> filterExpr: (key1 is not null and value1 is not null)
> (type: boolean)
> Statistics: Num rows: 20 Data size: 3560 Basic stats:
> PARTIAL Column stats: PARTIAL
> Filter Operator
> predicate: (key1 is not null and value1 is not null)
> (type: boolean)
> Statistics: Num rows: 20 Data size: 3560 Basic stats:
> PARTIAL Column stats: PARTIAL
> Select Operator
> expressions: key1 (type: string), value1 (type: string)
> outputColumnNames: _col0, _col1
> Statistics: Num rows: 20 Data size: 3560 Basic stats:
> PARTIAL Column stats: PARTIAL
> Reduce Output Operator
> key expressions: _col0 (type: string), _col1 (type:
> string)
> sort order: ++
> Map-reduce partition columns: _col0 (type: string),
> _col1 (type: string)
> Statistics: Num rows: 20 Data size: 3560 Basic stats:
> PARTIAL Column stats: PARTIAL
> Select Operator
> expressions: _col0 (type: string)
> outputColumnNames: _col0
> Statistics: Num rows: 20 Data size: 3560 Basic stats:
> PARTIAL Column stats: PARTIAL
> Group By Operator
> aggregations: min(_col0), max(_col0),
> bloom_filter(_col0, expectedEntries=20)
> mode: hash
> outputColumnNames: _col0, _col1, _col2
> Statistics: Num rows: 1 Data size: 730 Basic stats:
> PARTIAL Column stats: PARTIAL
> Reduce Output Operator
> sort order:
> Statistics: Num rows: 1 Data size: 730 Basic
> stats: PARTIAL Column stats: PARTIAL
> value expressions: _col0 (type: string), _col1
> (type: string), _col2 (type: binary)
> Execution mode: vectorized, llap
> LLAP IO: all inputs
> Reducer 2
> Execution mode: llap
> Reduce Operator Tree:
> Merge Join Operator
> condition map:
> Inner Join 0 to 1
> keys:
> 0 _col0 (type: string), _col1 (type: string)
> 1 _col0 (type: string), _col1 (type: string)
> Statistics: Num rows: 2200 Data size: 391600 Basic stats:
> PARTIAL Column stats: NONE
> Group By Operator
> aggregations: count()
> mode: hash
> outputColumnNames: _col0
> Statistics: Num rows: 1 Data size: 8 Basic stats: PARTIAL
> Column stats: NONE
> Reduce Output Operator
> sort order:
> Statistics: Num rows: 1 Data size: 8 Basic stats: PARTIAL
> Column stats: NONE
> value expressions: _col0 (type: bigint)
> Reducer 3
> Execution mode: vectorized, llap
> Reduce Operator Tree:
> Group By Operator
> aggregations: count(VALUE._col0)
> mode: mergepartial
> outputColumnNames: _col0
> Statistics: Num rows: 1 Data size: 8 Basic stats: PARTIAL
> Column stats: NONE
> File Output Operator
> compressed: false
> Statistics: Num rows: 1 Data size: 8 Basic stats: PARTIAL
> Column stats: NONE
> table:
> input format:
> org.apache.hadoop.mapred.SequenceFileInputFormat
> output format:
> org.apache.hadoop.hive.ql.io.HiveSequenceFileOutputFormat
> serde:
> org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe
> Reducer 5
> Execution mode: vectorized, llap
> Reduce Operator Tree:
> Group By Operator
> aggregations: min(VALUE._col0), max(VALUE._col1),
> bloom_filter(VALUE._col2, expectedEntries=20)
> mode: final
> outputColumnNames: _col0, _col1, _col2
> Statistics: Num rows: 1 Data size: 730 Basic stats: PARTIAL
> Column stats: PARTIAL
> Reduce Output Operator
> sort order:
> Statistics: Num rows: 1 Data size: 730 Basic stats: PARTIAL
> Column stats: PARTIAL
> value expressions: _col0 (type: string), _col1 (type:
> string), _col2 (type: binary)
> {code}
> Instead it should create one branch for a join with one bloom filter.
>
> The implementation for bloom filter requires getting a hash out of all the
> key columns and converting it to a long and feeding it to bloom filter as
> input. This requires a new UDF which does this. It will be called at both
> bloom filter generation and lookup phases.
> The min and max will stay independent as they are today for each columns.
> A vectorized implementation of such UDF is also required.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
