Implementation of count (*) in MySQL

stay MySQL in , Different storage engines ,count(*) There are different implementations .

MyISAM The engine stores the total row number of a table on disk , So execute count(*) It will return this number directly , It's very efficient

and InnoDB The engine is troublesome , It performs count(*) When , You need to read the data line by line from the engine , And then the cumulative count

All the above are without filter conditions count(*), If you add where conditions MyISAM Tables can't return so fast .

We also know that show table status If you order , It will be found that there is also a TABLE_ROWS Used to show how many rows this table currently has , This order is fast , So this one TABLE_ROWS Can replace count(*) Do you ？

answer ： Of course not ,TABLE_ROWS It is estimated by sampling , So it's not very accurate , Official documents mention that the error may reach 40%~50%, therefore show table status The number of lines displayed by the command cannot be used directly

Summary ：MyISAM Although the table count(*) Soon , But it doesn't support transactions

show table status The command returns quickly , But it's not accurate

InnoDB The table directly count(*) Will traverse the entire table , Although the results are accurate , But it will lead to performance problems

If you now have a page that often shows the total number of operation records of the trading system, what should you do ？

The answer is that we can only count by ourselves .

• Save the count with the cache system
• Save the count in the database

The way count(*)、count( Primary key id)、count( Field )、count(1) What are the differences in performance ？

• count( Primary key id),InnoDB The engine will traverse the entire table , Put each line of id Take out all the values , Return to server layer ,server Get the floor id after , Judgment cannot be empty , Just add up by lines .
• count(1),InnoDB The engine traverses the entire table , But no value ,server Layer for each row returned , Put a number “1” go in , Judgment cannot be empty , Add by line .

Just look at these two count(1) than count( Primary key id) Be quick , Because back from the engine id It's going to involve parsing the data lines , And copy field values .

• count( Field )
  • If this “ Field ” Is defined as not null Words , Read the field line by line from the record , Judgment cannot be null, Add by line ;
  • If this “ Field ” The definition is allowed to be null, So when it comes to execution , To judge that it might be null, We need to take out the value and judge again , No null Just add up .

                That's the first principle ,server What fields does the layer want ,InnoDB What fields are returned

• count(*)、 It doesn't take all the fields out , It's optimized , No value .count(*) Definitely not null, Add by line

Conclusion ： In order of efficiency ,count( Field ) < count( Primary key id) < count(1) About equal to count(*), So try to use count(*)