Help Center/ GaussDB/ Developer Guide(Centralized_8.x)/ SQL Optimization/ Optimization Cases/ Case: Creating an Appropriate Index
Updated on 2024-08-20 GMT+08:00
View PDF
Share

Case: Creating an Appropriate Index

Symptom

Query the information about all personnel in the sales department.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
 
-- Create a table.
CREATE TABLE staffs (staff_id NUMBER(6) NOT NULL, first_name VARCHAR2(20), last_name VARCHAR2(25), employment_id VARCHAR2(10), section_id NUMBER(4), state_name VARCHAR2(10), city VARCHAR2(10));
CREATE TABLE sections(section_id NUMBER(4), place_id NUMBER(4), section_name VARCHAR2(20));
CREATE TABLE states(state_id NUMBER(4));
CREATE TABLE places(place_id NUMBER(4), state_id  NUMBER(4));
-- Query before optimization.
EXPLAIN SELECT staff_id,first_name,last_name,employment_id,state_name,city 
FROM staffs,sections,states,places 
WHERE sections.section_name='Sales' 
AND staffs.section_id = sections.section_id 
AND sections.place_id = places.place_id 
AND places.state_id = states.state_id 
ORDER BY staff_id;
-- Query after optimization.
CREATE INDEX loc_id_pk ON places(place_id);
CREATE INDEX state_c_id_pk ON states(state_id);

EXPLAIN SELECT staff_id,first_name,last_name,employment_id,state_name,city 
FROM staffs,sections,states,places 
WHERE sections.section_name='Sales' 
AND staffs.section_id = sections.section_id 
AND sections.place_id = places.place_id 
AND places.state_id = states.state_id 
ORDER BY staff_id;

Optimization Analysis

The original execution plan is as follows before creating the places.place_id and states.state_id indexes:

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
 
                                             QUERY PLAN
----------------------------------------------------------------------------------------------------
 Sort  (cost=125.25..126.34 rows=438 width=254)
   Sort Key: staffs.staff_id
   ->  Hash Join  (cost=64.91..106.03 rows=438 width=254)
         Hash Cond: (states.state_id = places.state_id)
         ->  Seq Scan on states  (cost=0.00..29.45 rows=1945 width=12)
         ->  Hash  (cost=64.35..64.35 rows=45 width=266)
               ->  Hash Join  (cost=33.09..64.35 rows=45 width=266)
                     Hash Cond: (places.place_id = sections.place_id)
                     ->  Seq Scan on places  (cost=0.00..25.13 rows=1513 width=24)
                     ->  Hash  (cost=33.02..33.02 rows=6 width=266)
                           ->  Hash Join  (cost=19.16..33.02 rows=6 width=266)
                                 Hash Cond: (staffs.section_id = sections.section_id)
                                 ->  Seq Scan on staffs  (cost=0.00..12.76 rows=276 width=266)
                                 ->  Hash  (cost=19.11..19.11 rows=4 width=24)
                                       ->  Seq Scan on sections  (cost=0.00..19.11 rows=4 width=24)
                                             Filter: ((section_name)::text = 'Sales'::text)
(16 rows)

The optimized execution plan is as follows (two indexes have been created on the places.place_id and states.state_id columns):

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
 
                                             QUERY PLAN
-----------------------------------------------------------------------------------------------------
 Sort  (cost=107.40..108.49 rows=438 width=254)
   Sort Key: staffs.staff_id
   ->  Hash Join  (cost=35.37..88.18 rows=438 width=254)
         Hash Cond: (sections.section_id = staffs.section_id)
         ->  Nested Loop  (cost=19.16..66.85 rows=292 width=12)
               ->  Hash Join  (cost=19.16..50.27 rows=30 width=24)
                     Hash Cond: (places.place_id = sections.place_id)
                     ->  Seq Scan on places  (cost=0.00..25.13 rows=1513 width=24)
                     ->  Hash  (cost=19.11..19.11 rows=4 width=24)
                           ->  Seq Scan on sections  (cost=0.00..19.11 rows=4 width=24)
                                 Filter: ((section_name)::text = 'Sales'::text)
               ->  Index Only Scan using state_c_id_pk on states  (cost=0.00..0.45 rows=10 width=12)
                     Index Cond: (state_id = places.state_id)
         ->  Hash  (cost=12.76..12.76 rows=276 width=266)
               ->  Seq Scan on staffs  (cost=0.00..12.76 rows=276 width=266)
(15 rows)
Parent topic: Optimization Cases