1、查询优化应尽量避免全表扫描，考虑在where或order by条件字段上建立索引
~ 避免全表扫描的一些点：
1，where子句避免使用 != 或 <>;
2，where 子句避免中对字段进行 null 值判断，可以考虑字段非空设置，有默认值(0)
例：select id from tb where num is null
优：select id from t where num = 0
3，where 子句中应尽量避免使用 or 来连接条件
例：select id from tb where num=10 or num=20
优：select id from tb where num=10 union all select id from t where num=20
4，模糊查询尽量前边不使用%
例：select id from tb where name like '%abc%'
优：考虑全文搜索
5，慎用 in 和 not in，查询若连续考虑between ... and ...
例：select id from tb where num in(1,2,3)
优：select id from tb where num between 1 and 3
6，where 子句中避免尽量对字段进行表达式操作
例：select id from tb where num/2=100
优：select id from tb where num=100*2
7，where子句中避免尽量对字段进行函数操作
例：select id from tb where substring(name,1,3)='abc'
优：select id from tb where name like 'abc%'
例：select id from tb where datediff(day,createdate,'2005-11-30')=0--'2005-11-30'生成的id
优：select id from tb where createdate>='2005-11-30' and createdate<'2005-12-1'
待续。。。
