当前位置：网站首页>Sub query of database performance series

Sub query of database performance series

2022-07-27 10:15:00 【Dotnet cross platform】

Preface

Speaking of databases , Presumably some friends will think , Database is not every day CRUD Do you ？ As long as I master these tricks , It's nothing at all . Yes , In fact, I also agree with this view , For most applications , Only master these contents , Is competent for daily development work .

However, our program cannot always stay in the test environment , As the amount of production environment data increases , And the growth of user scale , There will be some performance problems in the database . So in the next period of time , We will learn some aspects of database performance optimization , I hope it can bring you some enlightenment and growth . today , Let's start with the most basic sub query .

summary

The definition of subquery itself is simple , Is a nested query . The most common example is that we have two tables students and classes, To simplify the description , Three paradigms are not followed here , The fields in the table are as follows ：

sutdents

id	stu_name	stu_age	class_id
1	Zhang San	15	1002
2	Li Si	15	1001
3	Wang Wu	16	1002
4	Zhu Liu	14	1003

classes

class_id	class_name
1001	Chinese language and literature
1002	mathematics
1003	English

At this point, we want to inquire which students have chosen math class , You can use subqueries to find results ：

SELECT stu_name FROM sutdents WHERE sutdents.class_id IN (SELECT class_id FROM classes WHERE class_name = ' mathematics ')

It can be seen that , We used... Here IN Operator to correlate the results of subqueries , Familiar partners must know , There is also a way of sub query called EXISTS, Use as follows ：

SELECT stu_name FROM sutdents s  WHERE EXISTS (SELECT 1 FROM classes c WHERE s.class_id = c.class_id AND class_name = ' mathematics ')

The query results of the two methods are consistent , We found out which students chose math ：

But actually , The efficiency of these two sub queries is different in different cases , Now let's talk about the difference between these two methods .

What is? IN

Literally ,IN That is, the fields of the external table , Use the results of internal nested queries as conditions to query the required results . therefore , Use IN when , Nested subqueries can only be queried once , Here comes a concept , be called ： Non-associative subqueries .

In our example above , The sub query will start from classes Find the course name in the table as “ mathematics ” Of id, At this time, the subquery is executed only once . Through the results of the query , That is, math class id Number , As the condition of external query , To find the final result .

What is? EXISTS

alike , seeing the name of a thing one thinks of its function ,EXISTS That is to say “ There is ”. in other words , When nesting queries , Information of external table , Need to exist in the result of internal table , Only then can the query result . therefore , Use EXISTS when , Nested subqueries will be queried multiple times , The number of times depends on the record entries of the external table , Finally, the corresponding result is returned to the external table , This concept is called ： Associated subquery .

In the example above , The query statement will start from students Find the corresponding class_id, Then pass it to the internal sub query , Internal subqueries use this class_id Match , Go to classes Search the table to meet class_id and class_name by “ mathematics ” The record of , Then return to the external table . In the process , Only exactly matched records , Will present the final result .

How to optimize subqueries

It is often mentioned in many previous experiences and some articles , Optimize SQL sentence , Need to use EXISTS Instead of IN, You can achieve the effect . So is this sentence correct ？

Through today's learning , We now know , That's not quite true , Not entirely wrong . The root cause is , We already know IN and EXISTS Characteristics of , So the conclusion is as follows ：

1、 If in a statement , Subquery external table A, The quantity is smaller than the table inside the sub query B, So at this time , Use EXISTS The efficiency of , Is greater than IN Of . This is because EXISTS Each query will cycle through the table A Data in , Then pass in the table B Match , In this case , Use EXISTS Fewer cycles , Fewer record entries need to be queried .

2、 conversely , If you subquery an external table A, The quantity is larger than the table inside the subquery B, So at this time , Use IN Will be more efficient . At this time to use IN You can narrow the range of records to be found , Final match table A Can reduce the size of records .

summary

Today we explain the concept of subquery and some simple usage , And some optimization methods . We can divide sub query into two modes: associated sub query and non associated sub query , They correspond to each other IN and EXISTS Usage of . Finally, when optimizing subqueries, we mentioned that which method to use depends on the size relationship between the outer table and the inner table , In fact, the focus of optimization is the principle of using small tables to drive large tables , Have you learned all these ？

In the next issue, we will continue to talk about how to optimize database performance from the perspective of database indexing , friends , See you next time ！