hive-commits mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From jd...@apache.org
Subject hive git commit: HIVE-17939: Bucket map join not being selected when bucketed tables is missing bucket files (Deepak Jaiswal, reviewed by Jason Dere)
Date Fri, 03 Nov 2017 18:06:07 GMT
Repository: hive
Updated Branches:
  refs/heads/master ef45f87f1 -> 56f5e5957


HIVE-17939: Bucket map join not being selected when bucketed tables is missing bucket files
(Deepak Jaiswal, reviewed by Jason Dere)


Project: http://git-wip-us.apache.org/repos/asf/hive/repo
Commit: http://git-wip-us.apache.org/repos/asf/hive/commit/56f5e595
Tree: http://git-wip-us.apache.org/repos/asf/hive/tree/56f5e595
Diff: http://git-wip-us.apache.org/repos/asf/hive/diff/56f5e595

Branch: refs/heads/master
Commit: 56f5e5957ea48404ae119721709d05600b59eeb1
Parents: ef45f87
Author: Jason Dere <jdere@hortonworks.com>
Authored: Fri Nov 3 11:05:44 2017 -0700
Committer: Jason Dere <jdere@hortonworks.com>
Committed: Fri Nov 3 11:05:44 2017 -0700

----------------------------------------------------------------------
 .../annotation/OpTraitsRulesProcFactory.java    |  47 +++----
 .../clientpositive/bucket_map_join_tez2.q       |  11 ++
 .../llap/bucket_map_join_tez2.q.out             | 130 ++++++++++++++++++
 .../spark/bucket_map_join_tez2.q.out            | 132 +++++++++++++++++++
 4 files changed, 298 insertions(+), 22 deletions(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/hive/blob/56f5e595/ql/src/java/org/apache/hadoop/hive/ql/optimizer/metainfo/annotation/OpTraitsRulesProcFactory.java
----------------------------------------------------------------------
diff --git a/ql/src/java/org/apache/hadoop/hive/ql/optimizer/metainfo/annotation/OpTraitsRulesProcFactory.java
b/ql/src/java/org/apache/hadoop/hive/ql/optimizer/metainfo/annotation/OpTraitsRulesProcFactory.java
index 191cfba..5933328 100644
--- a/ql/src/java/org/apache/hadoop/hive/ql/optimizer/metainfo/annotation/OpTraitsRulesProcFactory.java
+++ b/ql/src/java/org/apache/hadoop/hive/ql/optimizer/metainfo/annotation/OpTraitsRulesProcFactory.java
@@ -21,6 +21,7 @@ package org.apache.hadoop.hive.ql.optimizer.metainfo.annotation;
 import java.util.*;
 import java.util.Map.Entry;
 
+import org.apache.hadoop.hive.conf.HiveConf;
 import org.apache.hadoop.hive.metastore.api.Order;
 import org.apache.hadoop.hive.ql.exec.GroupByOperator;
 import org.apache.hadoop.hive.ql.exec.JoinOperator;
@@ -153,32 +154,34 @@ public class OpTraitsRulesProcFactory {
         return false;
       }
 
-      if (tbl.isPartitioned()) {
-        List<Partition> partitions = prunedParts.getNotDeniedPartns();
-        // construct a mapping of (Partition->bucket file names) and (Partition ->
bucket number)
-        if (!partitions.isEmpty()) {
-          for (Partition p : partitions) {
-            List<String> fileNames = 
-                AbstractBucketJoinProc.getBucketFilePathsOfPartition(p.getDataLocation(),

-                    pGraphContext);
-            // The number of files for the table should be same as number of
-            // buckets.
-            if (fileNames.size() != 0 && fileNames.size() != numBuckets) {
-              return false;
+      // Tez can handle unpopulated buckets
+      if (!HiveConf.getVar(pGraphContext.getConf(), HiveConf.ConfVars.HIVE_EXECUTION_ENGINE).equals("tez"))
{
+        if (tbl.isPartitioned()) {
+          List<Partition> partitions = prunedParts.getNotDeniedPartns();
+          // construct a mapping of (Partition->bucket file names) and (Partition ->
bucket number)
+          if (!partitions.isEmpty()) {
+            for (Partition p : partitions) {
+              List<String> fileNames =
+                      AbstractBucketJoinProc.getBucketFilePathsOfPartition(p.getDataLocation(),
+                              pGraphContext);
+              // The number of files for the table should be same as number of
+              // buckets.
+              if (fileNames.size() != 0 && fileNames.size() != numBuckets) {
+                return false;
+              }
             }
           }
-        }
-      } else {
-
-        List<String> fileNames =
-            AbstractBucketJoinProc.getBucketFilePathsOfPartition(tbl.getDataLocation(),
-                pGraphContext);
-        // The number of files for the table should be same as number of buckets.
-        if (fileNames.size() != 0 && fileNames.size() != numBuckets) {
-          return false;
+        } else {
+
+          List<String> fileNames =
+                  AbstractBucketJoinProc.getBucketFilePathsOfPartition(tbl.getDataLocation(),
+                          pGraphContext);
+          // The number of files for the table should be same as number of buckets.
+          if (fileNames.size() != 0 && fileNames.size() != numBuckets) {
+            return false;
+          }
         }
       }
-
       return true;
     }
 

http://git-wip-us.apache.org/repos/asf/hive/blob/56f5e595/ql/src/test/queries/clientpositive/bucket_map_join_tez2.q
----------------------------------------------------------------------
diff --git a/ql/src/test/queries/clientpositive/bucket_map_join_tez2.q b/ql/src/test/queries/clientpositive/bucket_map_join_tez2.q
index 45feec2..b79d89d 100644
--- a/ql/src/test/queries/clientpositive/bucket_map_join_tez2.q
+++ b/ql/src/test/queries/clientpositive/bucket_map_join_tez2.q
@@ -107,3 +107,14 @@ explain
 select count(*)
 from
 (select distinct key,value from tab_part) a join tab b on a.key = b.key and a.value = b.value;
+
+
+--HIVE-17939
+create table small (i int) stored as ORC;
+create table big (i int) partitioned by (k int) clustered by (i) into 10 buckets stored as
ORC;
+
+insert into small values (1),(2),(3),(4),(5),(6);
+insert into big partition(k=1) values(1),(3),(5),(7),(9);
+insert into big partition(k=2) values(0),(2),(4),(6),(8);
+explain select small.i, big.i from small,big where small.i=big.i;
+select small.i, big.i from small,big where small.i=big.i order by small.i, big.i;
\ No newline at end of file

http://git-wip-us.apache.org/repos/asf/hive/blob/56f5e595/ql/src/test/results/clientpositive/llap/bucket_map_join_tez2.q.out
----------------------------------------------------------------------
diff --git a/ql/src/test/results/clientpositive/llap/bucket_map_join_tez2.q.out b/ql/src/test/results/clientpositive/llap/bucket_map_join_tez2.q.out
index e2cee7f..b84edde 100644
--- a/ql/src/test/results/clientpositive/llap/bucket_map_join_tez2.q.out
+++ b/ql/src/test/results/clientpositive/llap/bucket_map_join_tez2.q.out
@@ -1707,3 +1707,133 @@ STAGE PLANS:
       Processor Tree:
         ListSink
 
+PREHOOK: query: create table small (i int) stored as ORC
+PREHOOK: type: CREATETABLE
+PREHOOK: Output: database:default
+PREHOOK: Output: default@small
+POSTHOOK: query: create table small (i int) stored as ORC
+POSTHOOK: type: CREATETABLE
+POSTHOOK: Output: database:default
+POSTHOOK: Output: default@small
+PREHOOK: query: create table big (i int) partitioned by (k int) clustered by (i) into 10
buckets stored as ORC
+PREHOOK: type: CREATETABLE
+PREHOOK: Output: database:default
+PREHOOK: Output: default@big
+POSTHOOK: query: create table big (i int) partitioned by (k int) clustered by (i) into 10
buckets stored as ORC
+POSTHOOK: type: CREATETABLE
+POSTHOOK: Output: database:default
+POSTHOOK: Output: default@big
+PREHOOK: query: insert into small values (1),(2),(3),(4),(5),(6)
+PREHOOK: type: QUERY
+PREHOOK: Output: default@small
+POSTHOOK: query: insert into small values (1),(2),(3),(4),(5),(6)
+POSTHOOK: type: QUERY
+POSTHOOK: Output: default@small
+POSTHOOK: Lineage: small.i EXPRESSION [(values__tmp__table__1)values__tmp__table__1.FieldSchema(name:tmp_values_col1,
type:string, comment:), ]
+PREHOOK: query: insert into big partition(k=1) values(1),(3),(5),(7),(9)
+PREHOOK: type: QUERY
+PREHOOK: Output: default@big@k=1
+POSTHOOK: query: insert into big partition(k=1) values(1),(3),(5),(7),(9)
+POSTHOOK: type: QUERY
+POSTHOOK: Output: default@big@k=1
+POSTHOOK: Lineage: big PARTITION(k=1).i EXPRESSION [(values__tmp__table__2)values__tmp__table__2.FieldSchema(name:tmp_values_col1,
type:string, comment:), ]
+PREHOOK: query: insert into big partition(k=2) values(0),(2),(4),(6),(8)
+PREHOOK: type: QUERY
+PREHOOK: Output: default@big@k=2
+POSTHOOK: query: insert into big partition(k=2) values(0),(2),(4),(6),(8)
+POSTHOOK: type: QUERY
+POSTHOOK: Output: default@big@k=2
+POSTHOOK: Lineage: big PARTITION(k=2).i EXPRESSION [(values__tmp__table__3)values__tmp__table__3.FieldSchema(name:tmp_values_col1,
type:string, comment:), ]
+PREHOOK: query: explain select small.i, big.i from small,big where small.i=big.i
+PREHOOK: type: QUERY
+POSTHOOK: query: explain select small.i, big.i from small,big where small.i=big.i
+POSTHOOK: type: QUERY
+STAGE DEPENDENCIES:
+  Stage-1 is a root stage
+  Stage-0 depends on stages: Stage-1
+
+STAGE PLANS:
+  Stage: Stage-1
+    Tez
+#### A masked pattern was here ####
+      Edges:
+        Map 2 <- Map 1 (CUSTOM_EDGE)
+#### A masked pattern was here ####
+      Vertices:
+        Map 1 
+            Map Operator Tree:
+                TableScan
+                  alias: small
+                  Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column stats:
NONE
+                  Filter Operator
+                    predicate: i is not null (type: boolean)
+                    Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column stats:
NONE
+                    Select Operator
+                      expressions: i (type: int)
+                      outputColumnNames: _col0
+                      Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column
stats: NONE
+                      Reduce Output Operator
+                        key expressions: _col0 (type: int)
+                        sort order: +
+                        Map-reduce partition columns: _col0 (type: int)
+                        Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column
stats: NONE
+            Execution mode: llap
+            LLAP IO: all inputs
+        Map 2 
+            Map Operator Tree:
+                TableScan
+                  alias: big
+                  Statistics: Num rows: 10 Data size: 80 Basic stats: COMPLETE Column stats:
NONE
+                  Filter Operator
+                    predicate: i is not null (type: boolean)
+                    Statistics: Num rows: 10 Data size: 80 Basic stats: COMPLETE Column stats:
NONE
+                    Select Operator
+                      expressions: i (type: int)
+                      outputColumnNames: _col0
+                      Statistics: Num rows: 10 Data size: 80 Basic stats: COMPLETE Column
stats: NONE
+                      Map Join Operator
+                        condition map:
+                             Inner Join 0 to 1
+                        keys:
+                          0 _col0 (type: int)
+                          1 _col0 (type: int)
+                        outputColumnNames: _col0, _col1
+                        input vertices:
+                          0 Map 1
+                        Statistics: Num rows: 11 Data size: 88 Basic stats: COMPLETE Column
stats: NONE
+                        File Output Operator
+                          compressed: false
+                          Statistics: Num rows: 11 Data size: 88 Basic stats: COMPLETE 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
+            Execution mode: llap
+            LLAP IO: all inputs
+
+  Stage: Stage-0
+    Fetch Operator
+      limit: -1
+      Processor Tree:
+        ListSink
+
+PREHOOK: query: select small.i, big.i from small,big where small.i=big.i order by small.i,
big.i
+PREHOOK: type: QUERY
+PREHOOK: Input: default@big
+PREHOOK: Input: default@big@k=1
+PREHOOK: Input: default@big@k=2
+PREHOOK: Input: default@small
+#### A masked pattern was here ####
+POSTHOOK: query: select small.i, big.i from small,big where small.i=big.i order by small.i,
big.i
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@big
+POSTHOOK: Input: default@big@k=1
+POSTHOOK: Input: default@big@k=2
+POSTHOOK: Input: default@small
+#### A masked pattern was here ####
+1	1
+2	2
+3	3
+4	4
+5	5
+6	6

http://git-wip-us.apache.org/repos/asf/hive/blob/56f5e595/ql/src/test/results/clientpositive/spark/bucket_map_join_tez2.q.out
----------------------------------------------------------------------
diff --git a/ql/src/test/results/clientpositive/spark/bucket_map_join_tez2.q.out b/ql/src/test/results/clientpositive/spark/bucket_map_join_tez2.q.out
index 34837f5..bd0f938 100644
--- a/ql/src/test/results/clientpositive/spark/bucket_map_join_tez2.q.out
+++ b/ql/src/test/results/clientpositive/spark/bucket_map_join_tez2.q.out
@@ -1636,3 +1636,135 @@ STAGE PLANS:
       Processor Tree:
         ListSink
 
+PREHOOK: query: create table small (i int) stored as ORC
+PREHOOK: type: CREATETABLE
+PREHOOK: Output: database:default
+PREHOOK: Output: default@small
+POSTHOOK: query: create table small (i int) stored as ORC
+POSTHOOK: type: CREATETABLE
+POSTHOOK: Output: database:default
+POSTHOOK: Output: default@small
+PREHOOK: query: create table big (i int) partitioned by (k int) clustered by (i) into 10
buckets stored as ORC
+PREHOOK: type: CREATETABLE
+PREHOOK: Output: database:default
+PREHOOK: Output: default@big
+POSTHOOK: query: create table big (i int) partitioned by (k int) clustered by (i) into 10
buckets stored as ORC
+POSTHOOK: type: CREATETABLE
+POSTHOOK: Output: database:default
+POSTHOOK: Output: default@big
+PREHOOK: query: insert into small values (1),(2),(3),(4),(5),(6)
+PREHOOK: type: QUERY
+PREHOOK: Output: default@small
+POSTHOOK: query: insert into small values (1),(2),(3),(4),(5),(6)
+POSTHOOK: type: QUERY
+POSTHOOK: Output: default@small
+POSTHOOK: Lineage: small.i EXPRESSION [(values__tmp__table__1)values__tmp__table__1.FieldSchema(name:tmp_values_col1,
type:string, comment:), ]
+PREHOOK: query: insert into big partition(k=1) values(1),(3),(5),(7),(9)
+PREHOOK: type: QUERY
+PREHOOK: Output: default@big@k=1
+POSTHOOK: query: insert into big partition(k=1) values(1),(3),(5),(7),(9)
+POSTHOOK: type: QUERY
+POSTHOOK: Output: default@big@k=1
+POSTHOOK: Lineage: big PARTITION(k=1).i EXPRESSION [(values__tmp__table__2)values__tmp__table__2.FieldSchema(name:tmp_values_col1,
type:string, comment:), ]
+PREHOOK: query: insert into big partition(k=2) values(0),(2),(4),(6),(8)
+PREHOOK: type: QUERY
+PREHOOK: Output: default@big@k=2
+POSTHOOK: query: insert into big partition(k=2) values(0),(2),(4),(6),(8)
+POSTHOOK: type: QUERY
+POSTHOOK: Output: default@big@k=2
+POSTHOOK: Lineage: big PARTITION(k=2).i EXPRESSION [(values__tmp__table__3)values__tmp__table__3.FieldSchema(name:tmp_values_col1,
type:string, comment:), ]
+PREHOOK: query: explain select small.i, big.i from small,big where small.i=big.i
+PREHOOK: type: QUERY
+POSTHOOK: query: explain select small.i, big.i from small,big where small.i=big.i
+POSTHOOK: type: QUERY
+STAGE DEPENDENCIES:
+  Stage-2 is a root stage
+  Stage-1 depends on stages: Stage-2
+  Stage-0 depends on stages: Stage-1
+
+STAGE PLANS:
+  Stage: Stage-2
+    Spark
+#### A masked pattern was here ####
+      Vertices:
+        Map 1 
+            Map Operator Tree:
+                TableScan
+                  alias: small
+                  Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column stats:
NONE
+                  Filter Operator
+                    predicate: i is not null (type: boolean)
+                    Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column stats:
NONE
+                    Select Operator
+                      expressions: i (type: int)
+                      outputColumnNames: _col0
+                      Statistics: Num rows: 6 Data size: 24 Basic stats: COMPLETE Column
stats: NONE
+                      Spark HashTable Sink Operator
+                        keys:
+                          0 _col0 (type: int)
+                          1 _col0 (type: int)
+            Local Work:
+              Map Reduce Local Work
+
+  Stage: Stage-1
+    Spark
+#### A masked pattern was here ####
+      Vertices:
+        Map 2 
+            Map Operator Tree:
+                TableScan
+                  alias: big
+                  Statistics: Num rows: 10 Data size: 40 Basic stats: COMPLETE Column stats:
NONE
+                  Filter Operator
+                    predicate: i is not null (type: boolean)
+                    Statistics: Num rows: 10 Data size: 40 Basic stats: COMPLETE Column stats:
NONE
+                    Select Operator
+                      expressions: i (type: int)
+                      outputColumnNames: _col0
+                      Statistics: Num rows: 10 Data size: 40 Basic stats: COMPLETE Column
stats: NONE
+                      Map Join Operator
+                        condition map:
+                             Inner Join 0 to 1
+                        keys:
+                          0 _col0 (type: int)
+                          1 _col0 (type: int)
+                        outputColumnNames: _col0, _col1
+                        input vertices:
+                          0 Map 1
+                        Statistics: Num rows: 11 Data size: 44 Basic stats: COMPLETE Column
stats: NONE
+                        File Output Operator
+                          compressed: false
+                          Statistics: Num rows: 11 Data size: 44 Basic stats: COMPLETE 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
+            Local Work:
+              Map Reduce Local Work
+
+  Stage: Stage-0
+    Fetch Operator
+      limit: -1
+      Processor Tree:
+        ListSink
+
+PREHOOK: query: select small.i, big.i from small,big where small.i=big.i order by small.i,
big.i
+PREHOOK: type: QUERY
+PREHOOK: Input: default@big
+PREHOOK: Input: default@big@k=1
+PREHOOK: Input: default@big@k=2
+PREHOOK: Input: default@small
+#### A masked pattern was here ####
+POSTHOOK: query: select small.i, big.i from small,big where small.i=big.i order by small.i,
big.i
+POSTHOOK: type: QUERY
+POSTHOOK: Input: default@big
+POSTHOOK: Input: default@big@k=1
+POSTHOOK: Input: default@big@k=2
+POSTHOOK: Input: default@small
+#### A masked pattern was here ####
+1	1
+2	2
+3	3
+4	4
+5	5
+6	6


Mime
View raw message