opt: add testcases showing stats problem

Release note: None
cockroachdb · Aug 21, 2018 · e1401e9 · e1401e9
1 parent 7855663
commit e1401e9
Show file tree

Hide file tree

Showing 2 changed files with 147 additions and 0 deletions.
diff --git a/pkg/sql/opt/memo/testdata/stats/join b/pkg/sql/opt/memo/testdata/stats/join
@@ -689,3 +689,83 @@ inner-join (merge)
       ├── right ordering: +5
       └── filters [type=bool, outer=(1,5), constraints=(/1: (/NULL - ]; /5: (/NULL - ]), fd=(1)==(5), (5)==(1)]
            └── a = e [type=bool, outer=(1,5), constraints=(/1: (/NULL - ]; /5: (/NULL - ])]
+
+exec-ddl
+CREATE TABLE uvw (u INT, v INT, w INT)
+----
+TABLE uvw
+ ├── u int
+ ├── v int
+ ├── w int
+ ├── rowid int not null (hidden)
+ └── INDEX primary
+      └── rowid int not null (hidden)
+
+exec-ddl
+CREATE TABLE xyz (x INT, y INT, z INT)
+----
+TABLE xyz
+ ├── x int
+ ├── y int
+ ├── z int
+ ├── rowid int not null (hidden)
+ └── INDEX primary
+      └── rowid int not null (hidden)
+
+# Verify that two equivalent formulations of a join lead to similar statistics.
+# In the first case, x=10 is pushed down; in the second case it is part of the
+# ON condition. The latter formulation happens in practice when we convert to
+# lookup join (we incorporate the filter back into the ON condition).
+# TODO(radu): the second case has orders of magnitude smaller row count; fix
+# this.
+
+norm disable=(PushFilterIntoJoinLeftAndRight,PushFilterIntoJoinLeft,PushFilterIntoJoinRight,MapFilterIntoJoinLeft,MapFilterIntoJoinRight)
+SELECT * FROM (SELECT * FROM uvw WHERE w=1) JOIN (SELECT * FROM xyz WHERE x=10) ON u=x
+----
+inner-join
+ ├── columns: u:1(int!null) v:2(int) w:3(int!null) x:5(int!null) y:6(int) z:7(int)
+ ├── stats: [rows=1.429009, distinct(1)=1, distinct(5)=1]
+ ├── fd: ()-->(1,3,5), (1)==(5), (5)==(1)
+ ├── select
+ │    ├── columns: u:1(int) v:2(int) w:3(int!null)
+ │    ├── stats: [rows=1.42857143, distinct(1)=1.42813399, distinct(3)=1]
+ │    ├── fd: ()-->(3)
+ │    ├── scan uvw
+ │    │    ├── columns: u:1(int) v:2(int) w:3(int)
+ │    │    └── stats: [rows=1000, distinct(1)=700, distinct(3)=700]
+ │    └── filters [type=bool, outer=(3), constraints=(/3: [/1 - /1]; tight), fd=()-->(3)]
+ │         └── w = 1 [type=bool, outer=(3), constraints=(/3: [/1 - /1]; tight)]
+ ├── select
+ │    ├── columns: x:5(int!null) y:6(int) z:7(int)
+ │    ├── stats: [rows=1.42857143, distinct(5)=1]
+ │    ├── fd: ()-->(5)
+ │    ├── scan xyz
+ │    │    ├── columns: x:5(int) y:6(int) z:7(int)
+ │    │    └── stats: [rows=1000, distinct(5)=700]
+ │    └── filters [type=bool, outer=(5), constraints=(/5: [/10 - /10]; tight), fd=()-->(5)]
+ │         └── x = 10 [type=bool, outer=(5), constraints=(/5: [/10 - /10]; tight)]
+ └── filters [type=bool, outer=(1,5), constraints=(/1: (/NULL - ]; /5: (/NULL - ]), fd=(1)==(5), (5)==(1)]
+      └── u = x [type=bool, outer=(1,5), constraints=(/1: (/NULL - ]; /5: (/NULL - ])]
+
+norm disable=(PushFilterIntoJoinLeftAndRight,PushFilterIntoJoinLeft,PushFilterIntoJoinRight,MapFilterIntoJoinLeft,MapFilterIntoJoinRight)
+SELECT * FROM (SELECT * FROM uvw WHERE w=1) JOIN xyz ON u=x AND x=10
+----
+inner-join
+ ├── columns: u:1(int!null) v:2(int) w:3(int!null) x:5(int!null) y:6(int) z:7(int)
+ ├── stats: [rows=0.0029154519, distinct(1)=0.0029154519, distinct(5)=0.0029154519]
+ ├── fd: ()-->(1,3,5), (1)==(5), (5)==(1)
+ ├── select
+ │    ├── columns: u:1(int) v:2(int) w:3(int!null)
+ │    ├── stats: [rows=1.42857143, distinct(1)=1.42813399, distinct(3)=1]
+ │    ├── fd: ()-->(3)
+ │    ├── scan uvw
+ │    │    ├── columns: u:1(int) v:2(int) w:3(int)
+ │    │    └── stats: [rows=1000, distinct(1)=700, distinct(3)=700]
+ │    └── filters [type=bool, outer=(3), constraints=(/3: [/1 - /1]; tight), fd=()-->(3)]
+ │         └── w = 1 [type=bool, outer=(3), constraints=(/3: [/1 - /1]; tight)]
+ ├── scan xyz
+ │    ├── columns: x:5(int) y:6(int) z:7(int)
+ │    └── stats: [rows=1000, distinct(5)=700]
+ └── filters [type=bool, outer=(1,5), constraints=(/1: (/NULL - ]; /5: [/10 - /10]), fd=()-->(1,5), (1)==(5), (5)==(1)]
+      ├── u = x [type=bool, outer=(1,5), constraints=(/1: (/NULL - ]; /5: (/NULL - ])]
+      └── x = 10 [type=bool, outer=(5), constraints=(/5: [/10 - /10]; tight)]
diff --git a/pkg/sql/opt/memo/testdata/stats/select b/pkg/sql/opt/memo/testdata/stats/select
@@ -588,3 +588,70 @@ select
  └── filters [type=bool, outer=(1), constraints=(/1: [/0 - /99]; tight)]
       ├── x >= 0 [type=bool, outer=(1), constraints=(/1: [/0 - ]; tight)]
       └── x < 100 [type=bool, outer=(1), constraints=(/1: (/NULL - /99]; tight)]
+
+exec-ddl
+CREATE TABLE uvw (u INT, v INT, w INT)
+----
+TABLE uvw
+ ├── u int
+ ├── v int
+ ├── w int
+ ├── rowid int not null (hidden)
+ └── INDEX primary
+      └── rowid int not null (hidden)
+
+# Test selectivity calculations by applying the two constraints in different
+# orders.
+# TODO(radu): applying both constraints at the same time results in much lower
+# estimates.
+norm
+SELECT * FROM uvw WHERE u=v AND u=10
+----
+select
+ ├── columns: u:1(int!null) v:2(int!null) w:3(int)
+ ├── stats: [rows=0.00204081633, distinct(1)=0.00204081633, distinct(2)=0.00204081633]
+ ├── fd: ()-->(1,2), (1)==(2), (2)==(1)
+ ├── scan uvw
+ │    ├── columns: u:1(int) v:2(int) w:3(int)
+ │    └── stats: [rows=1000, distinct(1)=700, distinct(2)=700]
+ └── filters [type=bool, outer=(1,2), constraints=(/1: [/10 - /10]; /2: (/NULL - ]), fd=()-->(1,2), (1)==(2), (2)==(1)]
+      ├── u = v [type=bool, outer=(1,2), constraints=(/1: (/NULL - ]; /2: (/NULL - ])]
+      └── u = 10 [type=bool, outer=(1), constraints=(/1: [/10 - /10]; tight)]
+
+norm disable=MergeSelects
+SELECT * FROM (SELECT * FROM uvw WHERE u=10) WHERE u=v
+----
+select
+ ├── columns: u:1(int!null) v:2(int!null) w:3(int)
+ ├── stats: [rows=1.0003063, distinct(1)=1, distinct(2)=1]
+ ├── fd: ()-->(1,2), (1)==(2), (2)==(1)
+ ├── select
+ │    ├── columns: u:1(int!null) v:2(int) w:3(int)
+ │    ├── stats: [rows=1.42857143, distinct(1)=1, distinct(2)=1.42813399]
+ │    ├── fd: ()-->(1)
+ │    ├── scan uvw
+ │    │    ├── columns: u:1(int) v:2(int) w:3(int)
+ │    │    └── stats: [rows=1000, distinct(1)=700, distinct(2)=700]
+ │    └── filters [type=bool, outer=(1), constraints=(/1: [/10 - /10]; tight), fd=()-->(1)]
+ │         └── u = 10 [type=bool, outer=(1), constraints=(/1: [/10 - /10]; tight)]
+ └── filters [type=bool, outer=(1,2), constraints=(/1: (/NULL - ]; /2: (/NULL - ]), fd=(1)==(2), (2)==(1)]
+      └── u = v [type=bool, outer=(1,2), constraints=(/1: (/NULL - ]; /2: (/NULL - ])]
+
+norm disable=MergeSelects
+SELECT * FROM (SELECT * FROM uvw WHERE u=v) WHERE u=10
+----
+select
+ ├── columns: u:1(int!null) v:2(int!null) w:3(int)
+ ├── stats: [rows=1, distinct(1)=1]
+ ├── fd: ()-->(1,2), (1)==(2), (2)==(1)
+ ├── select
+ │    ├── columns: u:1(int!null) v:2(int!null) w:3(int)
+ │    ├── stats: [rows=1.42857143, distinct(1)=1.42857143, distinct(2)=1.42857143]
+ │    ├── fd: (1)==(2), (2)==(1)
+ │    ├── scan uvw
+ │    │    ├── columns: u:1(int) v:2(int) w:3(int)
+ │    │    └── stats: [rows=1000, distinct(1)=700, distinct(2)=700]
+ │    └── filters [type=bool, outer=(1,2), constraints=(/1: (/NULL - ]; /2: (/NULL - ]), fd=(1)==(2), (2)==(1)]
+ │         └── u = v [type=bool, outer=(1,2), constraints=(/1: (/NULL - ]; /2: (/NULL - ])]
+ └── filters [type=bool, outer=(1), constraints=(/1: [/10 - /10]; tight), fd=()-->(1)]
+      └── u = 10 [type=bool, outer=(1), constraints=(/1: [/10 - /10]; tight)]