Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

indexrec: improve recommendations for queries with inequalities #104208

Merged
merged 1 commit into from
Jun 1, 2023
+218 −1
Merged

indexrec: improve recommendations for queries with inequalities #104208

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 6 additions & 1 deletion pkg/sql/opt/indexrec/candidate.go
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -29,7 +29,7 @@ import (
// indexes by table. For Order By, the index column ordering and column
// directions are the same as how it is in the Order By.
// 2. Add a single-column index on any Range expression, comparison
// expression (=, <, >, <=, >=), and IS expression.
// expression (=, !=, <, >, <=, >=), IS and IS NOT expression.
// 3. Add a single-column index on any column that appears in a JOIN predicate.
// 4. If there exist multiple columns from the same table in a JOIN predicate,
// create a single index on all such columns.
Expand Down Expand Up @@ -123,8 +123,13 @@ func (ics *indexCandidateSet) categorizeIndexCandidates(expr opt.Expr) {
case *memo.EqExpr:
ics.addVariableExprIndex(expr.Left, ics.equalCandidates)
ics.addVariableExprIndex(expr.Right, ics.equalCandidates)
case *memo.NeExpr:
ics.addVariableExprIndex(expr.Left, ics.rangeCandidates)
ics.addVariableExprIndex(expr.Right, ics.rangeCandidates)
case *memo.IsExpr:
ics.addVariableExprIndex(expr.Left, ics.equalCandidates)
case *memo.IsNotExpr:
ics.addVariableExprIndex(expr.Left, ics.rangeCandidates)
case *memo.LtExpr:
ics.addVariableExprIndex(expr.Left, ics.rangeCandidates)
ics.addVariableExprIndex(expr.Right, ics.rangeCandidates)
Expand Down
212 changes: 212 additions & 0 deletions pkg/sql/opt/indexrec/testdata/index
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -275,6 +275,42 @@ scan t1@_hyp_1
├── cost: 28.6200001
└── fd: ()-->(4)

index-candidates
SELECT i FROM t1 WHERE i != 5
----
t1:
(i)

index-recommendations
SELECT i FROM t1 WHERE i != 5
----
creation: CREATE INDEX ON t.public.t1 (i);
--
optimal plan:
scan t1@_hyp_1
├── columns: i:2!null
├── constraint: /2/5
│ ├── (/NULL - /4]
│ └── [/6 - ]
└── cost: 375.353333

index-candidates
SELECT s FROM t1 WHERE s IS NOT NULL
----
t1:
(s)

index-recommendations
SELECT s FROM t1 WHERE s IS NOT NULL
----
creation: CREATE INDEX ON t.public.t1 (s);
--
optimal plan:
scan t1@_hyp_1
├── columns: s:4!null
├── constraint: /4/5: (/NULL - ]
└── cost: 1067.42

index-candidates
SELECT t1.k FROM t1 JOIN t2 ON t1.k = t2.i
----
Expand Down Expand Up @@ -624,6 +660,32 @@ scan t1@_hyp_3
├── cost: 28.1933333
└── fd: ()-->(1)

index-candidates
SELECT * FROM t1 WHERE k = 1 AND f != 0
----
t1:
(f)
(k)
(k, f)

index-recommendations
SELECT * FROM t1 WHERE k = 1 AND f != 0
----
creation: CREATE INDEX ON t.public.t1 (k) STORING (i, f, s);
--
optimal plan:
select
├── columns: k:1!null i:2 f:3!null s:4
├── cost: 29.0500001
├── fd: ()-->(1)
├── scan t1@_hyp_1
│ ├── columns: k:1!null i:2 f:3 s:4
│ ├── constraint: /1/5: [/1 - /1]
│ ├── cost: 28.9200001
│ └── fd: ()-->(1)
└── filters
└── f:3 != 0.0 [outer=(3), constraints=(/3: (/NULL - /-5e-324] [/5e-324 - ]; tight)]

# Multi-column combinations used: EQ, EQ + R.
index-candidates
SELECT k, i, f FROM t1 WHERE k = 1 AND i = 2 AND f > 0
Expand Down Expand Up @@ -654,6 +716,28 @@ inner-join (zigzag t1@_hyp_1 t1@_hyp_2)
├── i:2 = 2 [outer=(2), constraints=(/2: [/2 - /2]; tight), fd=()-->(2)]
└── f:3 > 0.0 [outer=(3), constraints=(/3: [/5e-324 - ]; tight)]

index-candidates
SELECT k, i, s FROM t1 WHERE k = 1 AND i = 2 AND s IS NOT NULL
----
t1:
(i)
(k)
(k, i)
(k, i, s)
(s)

index-recommendations
SELECT k, i, s FROM t1 WHERE k = 1 AND i = 2 AND s IS NOT NULL
----
creation: CREATE INDEX ON t.public.t1 (k, i, s);
--
optimal plan:
scan t1@_hyp_5
├── columns: k:1!null i:2!null s:4!null
├── constraint: /1/2/4/5: (/1/2/NULL - /1/2]
├── cost: 18.907147
└── fd: ()-->(1,2)

# Multi-column combinations used: J + R.
index-candidates
SELECT t1.k, t1.f FROM t1 JOIN t2 ON t1.k != t2.k WHERE t1.f > 0
Expand Down Expand Up @@ -687,6 +771,39 @@ project
└── filters
└── t1.k:1 != t2.k:8 [outer=(1,8), constraints=(/1: (/NULL - ]; /8: (/NULL - ])]

index-candidates
SELECT t1.k, t1.s FROM t1 JOIN t2 ON t1.k = t2.k WHERE t1.s IS NOT NULL
----
t1:
(k)
(k, s)
(s)
t2:
(k)

index-recommendations
SELECT t1.k, t1.s FROM t1 JOIN t2 ON t1.k = t2.k WHERE t1.s IS NOT NULL
----
creation: CREATE INDEX ON t.public.t1 (s) STORING (k);
--
optimal plan:
project
├── columns: k:1!null s:4!null
├── cost: 2379.90804
└── inner-join (hash)
├── columns: t1.k:1!null t1.s:4!null t2.k:8!null
├── cost: 2282.85814
├── fd: (1)==(8), (8)==(1)
├── scan t2
│ ├── columns: t2.k:8
│ └── cost: 1078.52
├── scan t1@_hyp_1
│ ├── columns: t1.k:1 t1.s:4!null
│ ├── constraint: /4/5: (/NULL - ]
│ └── cost: 1077.32
└── filters
└── t1.k:1 = t2.k:8 [outer=(1,8), constraints=(/1: (/NULL - ]; /8: (/NULL - ]), fd=(1)==(8), (8)==(1)]

# Multi-column combinations used: EQ, EQ + J.
index-candidates
SELECT t1.i, t1.s FROM t1 JOIN t2 ON t1.k != t2.k WHERE t1.i = 2 AND t1.s = 'NG'
Expand Down Expand Up @@ -956,6 +1073,101 @@ union-all
└── aggregations
└── count-rows [as=count_rows:22]

index-candidates
SELECT count(*)
FROM t1 LEFT JOIN t2
ON t1.k = t2.k
GROUP BY t2.s, t2.i
UNION ALL
SELECT count(*)
FROM (
SELECT *
FROM t1
WHERE t1.f != 0
AND t1.s = 'NG'
)
----
t1:
(f)
(k)
(k, f)
(s)
(s, f)
(s, k)
(s, k, f)
t2:
(i, s)
(k)

index-recommendations
SELECT count(*)
FROM t1 LEFT JOIN t2
ON t1.k = t2.k
GROUP BY t2.s, t2.i
UNION ALL
SELECT count(*)
FROM (
SELECT *
FROM t1
WHERE t1.f != 0
AND t1.s = 'NG'
)
----
creation: CREATE INDEX ON t.public.t1 (k);
creation: CREATE INDEX ON t.public.t1 (s) STORING (f);
creation: CREATE INDEX ON t.public.t2 (k) STORING (i, s);
--
optimal plan:
union-all
├── columns: count:23!null
├── left columns: count_rows:14
├── right columns: count_rows:22
├── cardinality: [1 - ]
├── cost: 2766.55828
├── project
│ ├── columns: count_rows:14!null
│ ├── cost: 2727.55537
│ └── group-by (hash)
│ ├── columns: t2.i:9 t2.s:10 count_rows:14!null
│ ├── grouping columns: t2.i:9 t2.s:10
│ ├── cost: 2717.53581
│ ├── key: (9,10)
│ ├── fd: (9,10)-->(14)
│ ├── left-join (merge)
│ │ ├── columns: t1.k:1 t2.k:8 t2.i:9 t2.s:10
│ │ ├── left ordering: +1
│ │ ├── right ordering: +8
│ │ ├── cost: 2307.36
│ │ ├── scan t1@_hyp_3
│ │ │ ├── columns: t1.k:1
│ │ │ ├── cost: 1088.62
│ │ │ └── ordering: +1
│ │ ├── scan t2@_hyp_2
│ │ │ ├── columns: t2.k:8 t2.i:9 t2.s:10
│ │ │ ├── cost: 1098.72
│ │ │ └── ordering: +8
│ │ └── filters (true)
│ └── aggregations
│ └── count-rows [as=count_rows:14]
└── scalar-group-by
├── columns: count_rows:22!null
├── cardinality: [1 - 1]
├── cost: 28.9733334
├── key: ()
├── fd: ()-->(22)
├── select
│ ├── columns: f:17!null t1.s:18!null
│ ├── cost: 28.8500001
│ ├── fd: ()-->(18)
│ ├── scan t1@_hyp_1
│ │ ├── columns: f:17 t1.s:18!null
│ │ ├── constraint: /18/19: [/'NG' - /'NG']
│ │ ├── cost: 28.7200001
│ │ └── fd: ()-->(18)
│ └── filters
│ └── f:17 != 0.0 [outer=(17), constraints=(/17: (/NULL - /-5e-324] [/5e-324 - ]; tight)]
└── aggregations
└── count-rows [as=count_rows:22]

# No rule 5 multi-column index combinations.
index-candidates
Expand Down