记住不要出现查询order by，这样严重影响效率，当然除了最后可以出现
DATEDIFF是日期相减得到天数
DATE_FORMAT对日期格式化
DATEDIFF(DATE_FORMAT(NOW(),'%Y-%m-%d'), DATE_FORMAT(should_repay_date,'%Y-%m-%d')) AS should_repay_day
统计记录数使用count(1),不要用count(*)
select count(1) from user
查询字段不要用* 代替所有的，要查那些字段就查那些字段
select username, password from user
实现先排序后分组，不要用下面这个语句,外表查询每一条记录都会从内表遍历一遍，速度非常慢
select id, username, password, age from
(
select id, username, password, age from user order by age
) temp group by age
优化后
select t1.id, t1.usernam, t1.password, t1.age from user t1
inner join
(select id, max(age) as maxAge from user group by age ) b on t1.id = b.id and t1.age = b.maxAge