MySQL 45 learning notes (XI) how to index string fields

One . Prefix index

• MySQL Support prefix index , Sure Define part of the string Do the index
• Create an index statement Do not specify prefix length , that The index will contain the entire string

Two . How long prefix to use

Use prefix index , Define the length , Sure Save space , Don't add too many additional queries cost .

• Focus on the differentiation when building the index , The higher the discrimination, the better
• The more distinguishable , signify The fewer duplicate key values
• We can How many different values are there in the statistical index To determine how long prefix to use

Example ： Count the differentiation of different fields with different lengths

select 
  count(distinct left(email,4)）as L4,
  count(distinct left(email,5)）as L5,
  count(distinct left(email,6)）as L6,
  count(distinct left(email,7)）as L7,
from SUser;

3、 ... and . The influence of prefix index on overlay index

• If it is a full string , If the index coverage is supported, the table will not be returned .
• If it's just a part of a string , The system still depends on id Look back at the index , Because the system is not sure whether the prefix index definition truncates the complete information .

The longer the index , The more disk space it takes up , The smaller the index value that can be placed on the same data page , The less efficient the search .

Four . ID card scenario （ The front of the field should be highly differentiated from the back ）

• Use flashback storage , If you turn them upside down when storing ID number    
• Use hash Field

Use Flashback storage and Use hash Field The similarities and differences between the two methods

The same thing

• Range queries are not supported

Difference

• In terms of the extra space occupied , Flashbacks are stored in the primary key index , No additional storage is consumed
• hash Field method needs to add a field

CPU Consumption ：

• The flashback method needs an additional call reverse function
• hash Field requires an additional call to the hash function ,reverse Function extra consumption CPU Resources will be smaller

The query efficiency ：

• hash The query performance of field mode is relatively stable . because crc32 The calculated values have the probability of conflict , But the probability is very small
• After all, the reverse storage method is still the prefix index method , In other words, it will increase the number of scanning lines