[learning] database: several cases of index failure

Seven cases of index failure

Column and column comparison index failure

• Two columns are indexed separately , The following situations will not be indexed

select * from test where id = c_id

Column existence NULL Value index failure

• When we talk about database tables , It should be avoided as much as possible NULL Value appears
• If it can't be avoided , You'd better give one, too default The default value is
• The default value can be 0,-1
• The default value of the string can be “ An empty string ”

NOT Condition index failure

where The following conditions will cause the index to fail

• <>
• in
• not in
• not exists

select * from test where id<>500;
select * from test where id in (1,2,3,4,5);
select * from test where not in (6,7,8,9,0);
select * from test where not exists (select 1 from test_02 where test_02.id=test.id);

LIKE Wildcard index failure

• Try to use post wildcards for example ’name%’, Because when walking the index, it will match the index before , It can be found at this time
• If it is ‘%name’ Conditions of the query Pre wildcard
• The execution plan prefers to choose full table scanning , Therefore, the index will be invalid

The query condition includes function index invalidation

• In terms of query conditions, try not to use functions for indexes
• ABS,UPPER,DATE,DAY,YEAR etc.
• for example ：

select * from test where upper(name) = 'SUPPER';

select * from test where DATE(swap_time) = DATE( date_sub( CURRENT_DATE, INTERVAL 1 DAY ) );

• This is not going to be indexed , Because the index may be different when it is built and after calculation , Unable to locate index .
• But if the query condition is not to calculate the index column , So you can still go through the index , for example :

select * from test where name = upper('sunnnn');

Data type hermit conversion index failure

• When the query condition has hermit transformation , The index will fail
• For example, in the database id yes NUMBER type , But at the time of inquiry , But in the following form ：

select * from test where id = '123';

If msyql When you think full table scanning is faster （ Less data ） Index failure

When is index unnecessary ？

When the uniqueness of the field is poor

Frequently updated fields do not need to be

where Fields not used in conditions

Index usage <> when , Results the general