The execution efficiency of MySQL random fetching data

I found a problem last night when I was writing a random joke from a "passing" application, my goal was to randomly draw a joke from the joke table (about tens of thousands of data volumes) and show that my SQL was initially as follows:

SQL code

select　*　from　xiaohua　limit　随机数,1

Where random numbers are generated by Java programs, when the random number of hours, the discovery of the execution speed is very fast, but when the random number of more than 1w, the execution speed is very slow, it seems that the road is not workable, so I also use the rand () MySQL () function to see the implementation efficiency, and found that efficiency is not good, Finally, through Google, I found that there are many people encounter this problem, the final SQL is as follows (speed exceeded), can go to the http://l.faqee.com/to see the performance of the joke speed:

SQL code

SELECT　t1.id　as　id,title,content,class_name,class_id,hits　FROM　xiaohua　AS　t1　JOIN　(SELECT　ROUND(RAND()　*　((SELECT　MAX(id)　FROM　xiaohua　)-(SELECT　MIN(id)　FROM　xiaohua　))+(SELECT　MIN(id)　FROM　xiaohua　))　AS　id)　AS　t2　WHERE　t1.id　>=　t2.id　ORDER　BY　t1.id　LIMIT　1;

Forget about it!