HIVE-8978: Fix test determinism issue for qfile: smb_mapjoin_1.q etc …

…(Jimmy via Xuefu, reviewed by Brock) git-svn-id: https://svn.apache.org/repos/asf/hive/trunk@1642127 13f79535-47bb-0310-9956-ffa450edef68
gautamboddeda · Nov 27, 2014 · fc4a314 · fc4a314
1 parent 2e93b0c
commit fc4a314
Show file tree

Hide file tree

Showing 22 changed files with 318 additions and 175 deletions.
diff --git a/ql/src/test/queries/clientpositive/join_reorder.q b/ql/src/test/queries/clientpositive/join_reorder.q
@@ -10,6 +10,8 @@ LOAD DATA LOCAL INPATH '../../data/files/T1.txt' INTO TABLE T1;
 LOAD DATA LOCAL INPATH '../../data/files/T2.txt' INTO TABLE T2;
 LOAD DATA LOCAL INPATH '../../data/files/T3.txt' INTO TABLE T3;
 
+-- SORT_QUERY_RESULTS
+
 EXPLAIN FROM T1 a JOIN src c ON c.key+1=a.key
 SELECT a.key, a.val, c.key;
 

diff --git a/ql/src/test/queries/clientpositive/outer_join_ppr.q b/ql/src/test/queries/clientpositive/outer_join_ppr.q
@@ -1,5 +1,7 @@
 set hive.optimize.ppd=true;
 
+-- SORT_QUERY_RESULTS
+
 EXPLAIN EXTENDED
  FROM 
   src a

diff --git a/ql/src/test/queries/clientpositive/smb_mapjoin_1.q b/ql/src/test/queries/clientpositive/smb_mapjoin_1.q
@@ -13,7 +13,9 @@ load data local inpath '../../data/files/smbbucket_3.rc' overwrite into table sm
 set hive.optimize.bucketmapjoin = true;
 set hive.optimize.bucketmapjoin.sortedmerge = true;
 set hive.input.format = org.apache.hadoop.hive.ql.io.BucketizedHiveInputFormat;
-
+
+-- SORT_QUERY_RESULTS
+
 explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key;
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key;

diff --git a/ql/src/test/queries/clientpositive/smb_mapjoin_2.q b/ql/src/test/queries/clientpositive/smb_mapjoin_2.q
@@ -13,7 +13,9 @@ load data local inpath '../../data/files/smbbucket_3.rc' overwrite into table sm
 set hive.optimize.bucketmapjoin = true;
 set hive.optimize.bucketmapjoin.sortedmerge = true;
 set hive.input.format = org.apache.hadoop.hive.ql.io.BucketizedHiveInputFormat;
-
+
+-- SORT_QUERY_RESULTS
+
 explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_3 b on a.key = b.key;
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_3 b on a.key = b.key;

diff --git a/ql/src/test/queries/clientpositive/smb_mapjoin_4.q b/ql/src/test/queries/clientpositive/smb_mapjoin_4.q
@@ -13,7 +13,9 @@ load data local inpath '../../data/files/smbbucket_3.rc' overwrite into table sm
 set hive.optimize.bucketmapjoin = true;
 set hive.optimize.bucketmapjoin.sortedmerge = true;
 set hive.input.format = org.apache.hadoop.hive.ql.io.BucketizedHiveInputFormat;
-
+
+-- SORT_QUERY_RESULTS
+
 explain
 select /*+mapjoin(a,b)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key join smb_bucket_3 c on b.key=c.key;
 select /*+mapjoin(a,b)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key join smb_bucket_3 c on b.key=c.key;

diff --git a/ql/src/test/queries/clientpositive/smb_mapjoin_5.q b/ql/src/test/queries/clientpositive/smb_mapjoin_5.q
@@ -13,7 +13,9 @@ load data local inpath '../../data/files/smbbucket_3.rc' overwrite into table sm
 set hive.optimize.bucketmapjoin = true;
 set hive.optimize.bucketmapjoin.sortedmerge = true;
 set hive.input.format = org.apache.hadoop.hive.ql.io.BucketizedHiveInputFormat;
-
+
+-- SORT_QUERY_RESULTS
+
 explain
 select /*+mapjoin(a,c)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key join smb_bucket_3 c on b.key=c.key;
 select /*+mapjoin(a,c)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key join smb_bucket_3 c on b.key=c.key;

diff --git a/ql/src/test/queries/clientpositive/smb_mapjoin_8.q b/ql/src/test/queries/clientpositive/smb_mapjoin_8.q
@@ -3,6 +3,7 @@ set hive.enforce.bucketing = true;
 set hive.enforce.sorting = true;
 set hive.exec.reducers.max = 1;
 
+-- SORT_QUERY_RESULTS
 
 create table smb_bucket_input (key int, value string) stored as rcfile;
 load data local inpath '../../data/files/smb_bucket_input.rc' into table smb_bucket_input;

diff --git a/ql/src/test/queries/clientpositive/uniquejoin.q b/ql/src/test/queries/clientpositive/uniquejoin.q
@@ -6,6 +6,8 @@ LOAD DATA LOCAL INPATH '../../data/files/T1.txt' INTO TABLE T1;
 LOAD DATA LOCAL INPATH '../../data/files/T2.txt' INTO TABLE T2;
 LOAD DATA LOCAL INPATH '../../data/files/T3.txt' INTO TABLE T3;
 
+-- SORT_QUERY_RESULTS
+
 FROM UNIQUEJOIN PRESERVE T1 a (a.key), PRESERVE T2 b (b.key), PRESERVE T3 c (c.key)
 SELECT a.key, b.key, c.key;
 

diff --git a/ql/src/test/queries/clientpositive/vector_decimal_aggregate.q b/ql/src/test/queries/clientpositive/vector_decimal_aggregate.q
@@ -6,6 +6,8 @@ CREATE TABLE decimal_vgby STORED AS ORC AS
 
 SET hive.vectorized.execution.enabled=true;
 
+-- SORT_QUERY_RESULTS
+
 -- First only do simple aggregations that output primitives only
 EXPLAIN SELECT cint,
     COUNT(cdecimal1), MAX(cdecimal1), MIN(cdecimal1), SUM(cdecimal1),

diff --git a/ql/src/test/queries/clientpositive/vectorization_13.q b/ql/src/test/queries/clientpositive/vectorization_13.q
@@ -31,6 +31,7 @@ WHERE    (((cfloat < 3569)
               AND ((ctimestamp2 != 12)
                    AND (ctinyint < 9763215.5639))))
 GROUP BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
+ORDER BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
 LIMIT 40;
 
 SELECT   cboolean1,
@@ -62,6 +63,7 @@ WHERE    (((cfloat < 3569)
               AND ((ctimestamp2 != 12)
                    AND (ctinyint < 9763215.5639))))
 GROUP BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
+ORDER BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
 LIMIT 40;
 
 -- double compare timestamp
@@ -95,6 +97,7 @@ WHERE    (((cfloat < 3569)
               AND ((ctimestamp2 != -1.3359999999999999)
                    AND (ctinyint < 9763215.5639))))
 GROUP BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
+ORDER BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
 LIMIT 40;
 
 SELECT   cboolean1,
@@ -126,5 +129,6 @@ WHERE    (((cfloat < 3569)
               AND ((ctimestamp2 != -1.3359999999999999)
                    AND (ctinyint < 9763215.5639))))
 GROUP BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
+ORDER BY cboolean1, ctinyint, ctimestamp1, cfloat, cstring1
 LIMIT 40;
 
diff --git a/ql/src/test/results/clientpositive/join_reorder.q.out b/ql/src/test/results/clientpositive/join_reorder.q.out
@@ -46,10 +46,14 @@ POSTHOOK: query: LOAD DATA LOCAL INPATH '../../data/files/T3.txt' INTO TABLE T3
 POSTHOOK: type: LOAD
 #### A masked pattern was here ####
 POSTHOOK: Output: default@t3
-PREHOOK: query: EXPLAIN FROM T1 a JOIN src c ON c.key+1=a.key
+PREHOOK: query: -- SORT_QUERY_RESULTS
+
+EXPLAIN FROM T1 a JOIN src c ON c.key+1=a.key
 SELECT a.key, a.val, c.key
 PREHOOK: type: QUERY
-POSTHOOK: query: EXPLAIN FROM T1 a JOIN src c ON c.key+1=a.key
+POSTHOOK: query: -- SORT_QUERY_RESULTS
+
+EXPLAIN FROM T1 a JOIN src c ON c.key+1=a.key
 SELECT a.key, a.val, c.key
 POSTHOOK: type: QUERY
 STAGE DEPENDENCIES:
@@ -421,9 +425,9 @@ POSTHOOK: Input: default@t2
 POSTHOOK: Input: default@t3
 #### A masked pattern was here ####
 2	2	12	12
+7	NULL	17	17
 NULL	NULL	NULL	14
 NULL	NULL	NULL	16
-7	NULL	17	17
 PREHOOK: query: FROM T1 a
   LEFT OUTER JOIN T2 b ON (b.key=a.key)
   RIGHT OUTER JOIN T3 c ON (c.val = a.val)
@@ -443,9 +447,9 @@ POSTHOOK: Input: default@t2
 POSTHOOK: Input: default@t3
 #### A masked pattern was here ####
 2	2	12	12
+7	NULL	17	17
 NULL	NULL	NULL	14
 NULL	NULL	NULL	16
-7	NULL	17	17
 PREHOOK: query: EXPLAIN FROM UNIQUEJOIN
   PRESERVE T1 a (a.key, a.val),
   PRESERVE T2 b (b.key, b.val),
@@ -616,15 +620,15 @@ POSTHOOK: Input: default@t3
 #### A masked pattern was here ####
 1	NULL	NULL
 2	NULL	2
-NULL	2	NULL
 3	3	NULL
-NULL	4	4
-NULL	5	NULL
-NULL	NULL	6
 7	NULL	7
 8	8	NULL
 8	8	NULL
 8	NULL	NULL
+NULL	2	NULL
+NULL	4	4
+NULL	5	NULL
+NULL	NULL	6
 PREHOOK: query: FROM UNIQUEJOIN
   PRESERVE T1 a (a.key, a.val),
   PRESERVE T2 b (b.key, b.val),
@@ -647,12 +651,12 @@ POSTHOOK: Input: default@t3
 #### A masked pattern was here ####
 1	NULL	NULL
 2	NULL	2
-NULL	2	NULL
 3	3	NULL
-NULL	4	4
-NULL	5	NULL
-NULL	NULL	6
 7	NULL	7
 8	8	NULL
 8	8	NULL
 8	NULL	NULL
+NULL	2	NULL
+NULL	4	4
+NULL	5	NULL
+NULL	NULL	6
diff --git a/ql/src/test/results/clientpositive/outer_join_ppr.q.out b/ql/src/test/results/clientpositive/outer_join_ppr.q.out
diff --git a/ql/src/test/results/clientpositive/smb_mapjoin_1.q.out b/ql/src/test/results/clientpositive/smb_mapjoin_1.q.out
@@ -46,10 +46,14 @@ POSTHOOK: query: load data local inpath '../../data/files/smbbucket_3.rc' overwr
 POSTHOOK: type: LOAD
 #### A masked pattern was here ####
 POSTHOOK: Output: default@smb_bucket_3
-PREHOOK: query: explain
+PREHOOK: query: -- SORT_QUERY_RESULTS
+
+explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key
 PREHOOK: type: QUERY
-POSTHOOK: query: explain
+POSTHOOK: query: -- SORT_QUERY_RESULTS
+
+explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_2 b on a.key = b.key
 POSTHOOK: type: QUERY
 STAGE DEPENDENCIES:
@@ -156,10 +160,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_2
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	NULL	NULL
 3	val_3	NULL	NULL
 4	val_4	NULL	NULL
 5	val_5	NULL	NULL
-10	val_10	NULL	NULL
 PREHOOK: query: explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a right outer join smb_bucket_2 b on a.key = b.key
 PREHOOK: type: QUERY
@@ -271,10 +275,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_2
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	NULL	NULL
 3	val_3	NULL	NULL
 4	val_4	NULL	NULL
 5	val_5	NULL	NULL
-10	val_10	NULL	NULL
 NULL	NULL	20	val_20
 NULL	NULL	23	val_23
 NULL	NULL	25	val_25
@@ -389,10 +393,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_2
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	NULL	NULL
 3	val_3	NULL	NULL
 4	val_4	NULL	NULL
 5	val_5	NULL	NULL
-10	val_10	NULL	NULL
 PREHOOK: query: explain
 select /*+mapjoin(b)*/ * from smb_bucket_1 a right outer join smb_bucket_2 b on a.key = b.key
 PREHOOK: type: QUERY
@@ -504,10 +508,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_2
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	NULL	NULL
 3	val_3	NULL	NULL
 4	val_4	NULL	NULL
 5	val_5	NULL	NULL
-10	val_10	NULL	NULL
 NULL	NULL	20	val_20
 NULL	NULL	23	val_23
 NULL	NULL	25	val_25

diff --git a/ql/src/test/results/clientpositive/smb_mapjoin_2.q.out b/ql/src/test/results/clientpositive/smb_mapjoin_2.q.out
@@ -46,10 +46,14 @@ POSTHOOK: query: load data local inpath '../../data/files/smbbucket_3.rc' overwr
 POSTHOOK: type: LOAD
 #### A masked pattern was here ####
 POSTHOOK: Output: default@smb_bucket_3
-PREHOOK: query: explain
+PREHOOK: query: -- SORT_QUERY_RESULTS
+
+explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_3 b on a.key = b.key
 PREHOOK: type: QUERY
-POSTHOOK: query: explain
+POSTHOOK: query: -- SORT_QUERY_RESULTS
+
+explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a join smb_bucket_3 b on a.key = b.key
 POSTHOOK: type: QUERY
 STAGE DEPENDENCIES:
@@ -102,8 +106,8 @@ POSTHOOK: type: QUERY
 POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
-4	val_4	4	val_4
 10	val_10	10	val_10
+4	val_4	4	val_4
 PREHOOK: query: explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a left outer join smb_bucket_3 b on a.key = b.key
 PREHOOK: type: QUERY
@@ -158,10 +162,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	10	val_10
 3	val_3	NULL	NULL
 4	val_4	4	val_4
 5	val_5	NULL	NULL
-10	val_10	10	val_10
 PREHOOK: query: explain
 select /*+mapjoin(a)*/ * from smb_bucket_1 a right outer join smb_bucket_3 b on a.key = b.key
 PREHOOK: type: QUERY
@@ -215,8 +219,8 @@ POSTHOOK: type: QUERY
 POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
-4	val_4	4	val_4
 10	val_10	10	val_10
+4	val_4	4	val_4
 NULL	NULL	17	val_17
 NULL	NULL	19	val_19
 NULL	NULL	20	val_20
@@ -275,10 +279,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	10	val_10
 3	val_3	NULL	NULL
 4	val_4	4	val_4
 5	val_5	NULL	NULL
-10	val_10	10	val_10
 NULL	NULL	17	val_17
 NULL	NULL	19	val_19
 NULL	NULL	20	val_20
@@ -339,8 +343,8 @@ POSTHOOK: type: QUERY
 POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
-4	val_4	4	val_4
 10	val_10	10	val_10
+4	val_4	4	val_4
 PREHOOK: query: explain
 select /*+mapjoin(b)*/ * from smb_bucket_1 a left outer join smb_bucket_3 b on a.key = b.key
 PREHOOK: type: QUERY
@@ -395,10 +399,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	10	val_10
 3	val_3	NULL	NULL
 4	val_4	4	val_4
 5	val_5	NULL	NULL
-10	val_10	10	val_10
 PREHOOK: query: explain
 select /*+mapjoin(b)*/ * from smb_bucket_1 a right outer join smb_bucket_3 b on a.key = b.key
 PREHOOK: type: QUERY
@@ -452,8 +456,8 @@ POSTHOOK: type: QUERY
 POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
-4	val_4	4	val_4
 10	val_10	10	val_10
+4	val_4	4	val_4
 NULL	NULL	17	val_17
 NULL	NULL	19	val_19
 NULL	NULL	20	val_20
@@ -512,10 +516,10 @@ POSTHOOK: Input: default@smb_bucket_1
 POSTHOOK: Input: default@smb_bucket_3
 #### A masked pattern was here ####
 1	val_1	NULL	NULL
+10	val_10	10	val_10
 3	val_3	NULL	NULL
 4	val_4	4	val_4
 5	val_5	NULL	NULL
-10	val_10	10	val_10
 NULL	NULL	17	val_17
 NULL	NULL	19	val_19
 NULL	NULL	20	val_20