MySQL ORDER BY Rand () efficiency

Recently, because of the need to study the MySQL random extraction implementation method. For example, to randomly extract a record from the TableName table, the general wording would be: SELECT * from TableName ORDER by RAND () LIMIT 1.

But then I looked up the official MySQL manual, and the hint for rand () probably meant that the rand () function could not be used in an ORDER BY clause because it would cause data columns to be scanned multiple times. However, in the MySQL 3.23 version, it is still possible to implement random by the order by RAND ().

But the real test was found to be very inefficient. A library of more than 150,000, query 5 data, incredibly more than 8 seconds. View the official manual, also said that Rand () is executed multiple times in the ORDER BY clause, which is naturally inefficient and very low.

You are cannot use a column with RAND () of the clause, because order by would evaluate the column multiple time S.

Search Google, the Internet is basically query Max (ID) * RAND () to randomly 

SELECT *
FROM `table` AS t1 JOIN (SELECT ROUND(RAND() * (SELECT MAX(id) FROM `table`)) AS id) AS t2
WHERE t1.id >= t2.id
ORDER BY t1.id ASC LIMIT 5;

But this will produce a continuous 5 records. The solution can only be one query at a time, query 5 times. Even so, because of the 150,000 table, the query only needs 0.01 seconds.

The following statements are used by Join,mysql forums where someone uses the

SELECT *
FROM `table`
WHERE id >= (SELECT FLOOR( MAX(id) * RAND()) FROM `table` )
ORDER BY id LIMIT 1;

I tested it, it takes 0.5 seconds, and the speed is good, but there is still a big gap with the above statement. There is something wrong with the total sleep.

So I rewrote the statement.

SELECT * FROM `table`
WHERE id >= (SELECT floor(RAND() * (SELECT MAX(id) FROM `table`))) 
ORDER BY id LIMIT 1;

This, the efficiency is increased, the query time is only 0.01 seconds

Finally, the statement to improve, plus the min (id) judgment. I was at the beginning of the test, because I did not add min (id) judgment, the result is half of the time is always query to the first few lines in the table.

The full query statement is:

SELECT * FROM `table`
WHERE id >= (SELECT floor( RAND() * ((SELECT MAX(id) FROM `table`)-(SELECT MIN(id) FROM `table`)) + (SELECT MIN(id) FROM `table`))) 
ORDER BY id LIMIT 1;
  
SELECT *
FROM `table` AS t1 JOIN (SELECT ROUND(RAND() * ((SELECT MAX(id) FROM `table`)-(SELECT MIN(id) FROM `table`))+(SELECT MIN(id) FROM `table`)) AS id) AS t2
WHERE t1.id >= t2.id
ORDER BY t1.id LIMIT 1;

Finally in PHP, the two statements are queried separately 10 times,

The former takes 0.147433 seconds.

The latter takes time 0.015130 seconds

It seems that the syntax for join is much higher than the efficiency of using functions directly in a where.