Several types of MySQL index invalidation and their solutions

Click below “Java Programming duck ” Follow and mark the stars

More exciting First time direct

What are the conditions of index failure ？

1. Index columns are not independent

2. Left blur is used

3. Use or Some fields in the query do not use an index

4. String conditions are not used ''

5. Queries that do not conform to the leftmost prefix principle

6. Index field not added not null constraint

7. Implicit conversion leads to index invalidation

Index column independence means that the indexed column cannot be part of an expression , Cannot be an argument to a function , For example, the following situation

select id,name,age,salary from table_name where salary + 1000 = 6000;

salary The column is evaluated by the user expression , In this case, the index will fail , The solution is to calculate the condition value in advance , Don't let the index column participate in the evaluation .

The index field is used as the parameter of the function

select id,name,age,salary from table_name where substring(name,1,3)= 'luc';

What is the solution , You can calculate the conditions in advance , Do not use indexes , Or you can use other sql Replace the above , such as , above sql have access to like Instead of

select id,name,age,salary from table_name where name like 'luc%';

Left blur is used

select id,name,age,salary from table_name where name like '%lucs%';

Avoid using left blur as much as possible , It can be written like this

select id,name,age,salary from table_name where name like 'lucs%';

If you can't avoid left fuzzy query , Consider search engines such as ES

or Some fields in the query do not use an index

select id,name,age,salary from table_name where name ='lucs' and age >25

This situation , It can be for name and age All indexed , Otherwise, it will take a full scan .

String conditions are not used ''

select id,name,age,salary from table_name where phone=13088772233

The one above sql phone The field type is String type , But it's not used '13088772233 ', SQL Just scan the whole table , So the string index should use ‘’

select id,name,age,salary from table_name where phone='13088772233 '

Queries that do not conform to the leftmost prefix principle

For example, there is such a composite index index(a,b,c)

select * from table_name where b='1'and c='2'
select * from table_name where c='2'


//  The top two  SQL  Can't run the index 

The principle of left , Is to have the leftmost priority , In my absence , You can't play by yourself , Unless you create your own index , Here are a few SQL You can execute the index

select * from table_name where a = 'asaa' and b='1'and c='2'
select * from table_name where a = 'asda' and b='1231' 
//  The above two items are indexed , But how do you think the index should go , Yes, all go , Or partial indexing ？
select * from table_name where a = 'asda' and c='dsfsdafsfsd'

Index field not added not null constraint

select * from table_name where a is null;
//  This article sql You can't go index execution ,is null  Conditions   Index cannot be used , Can only scan the whole table 
// mysql  The official suggestion is to set the field to  not null

So in this case , stay mysql When creating table fields , You can set the string to be indexed to not null default '' The default is an empty string

Implicit conversion

Implicit conversion will occur if the two field types of the associated table are inconsistent

select * from table_name t1 left join table_name2 t2 on t1.id=t2.tid;
//  In the above sentence , If  t1  Tabular id  The type and  t2  Tabular tid  When the types are inconsistent , Can not 
//  Executed by index .
//  The solution is to set the field types uniformly .

END

 After reading this article, there are gains ？ Please forward to share with more people 
 Focus on 「Java Programming duck 」, promote Java Skill 
 Focus on Java Programming duck WeChat official account , The background to reply ： Yard farm gift bag   A copy of the latest technical data can be obtained . cover Java  Frame learning 、 Architect learning, etc ！


 If the article helps , Looking at , Forward! .
 Thank you for your support  (*^__^*）