评论：mysql distinct和group by谁更好 http://blog.51yip.com/mysql/1105.html －－一步，二步，三步，N步，二行脚印 Tue, 07 Jun 2022 01:26:20 +0000 http://wordpress.org/?v=2.9.1 hourly 1 来自：张映 http://blog.51yip.com/mysql/1105.html/comment-page-1#comment-349 张映 Fri, 26 Nov 2010 01:43:54 +0000 http://blog.51yip.com/?p=1105#comment-349 索引就好比是一本书的目录，如果这本书很厚的话，找起东西来比较麻烦的，比如汉语字典，找字的时候，我们会先去找目录，因为这样找比较快一点。如果一本书，只有十几页，这个时候，我们找东西，根本不要去看目录了，因为这样更浪费时间。 从你贴出来的数据看，你测试的数据表里面，数据量比较小，这个时候加索引反而会影响速度的。 索引就好比是一本书的目录，如果这本书很厚的话，找起东西来比较麻烦的，比如汉语字典，找字的时候，我们会先去找目录，因为这样找比较快一点。如果一本书，只有十几页，这个时候，我们找东西，根本不要去看目录了，因为这样更浪费时间。

从你贴出来的数据看，你测试的数据表里面，数据量比较小，这个时候加索引反而会影响速度的。

]]> 来自：kenny http://blog.51yip.com/mysql/1105.html/comment-page-1#comment-348 kenny Fri, 26 Nov 2010 01:22:40 +0000 http://blog.51yip.com/?p=1105#comment-348 存储过程 DELIMITER $$ DROP PROCEDURE IF EXISTS `test`.`dis`$$ CREATE DEFINER=`root`@`localhost` PROCEDURE `dis`(num int(10)) begin declare i int(10); set i = 1; while i <= num do insert into dis_group (num) values (floor(rand()*100)); set i=i+1; end while; end$$ DELIMITER ; | Query_ID | Duration | Query | +----------+------------+------------------------------------------+ | 1 | 0.00099400 | show variables like "%pro%" | | 2 | 0.09483400 | select distinct(num) from dis_group | | 3 | 0.09505400 | select num from dis_group group by num | | 4 | 0.00004600 | select num from dis_group group by num | | 5 | 0.00004200 | select distinct(num) from dis_group | | 6 | 0.00004400 | select distinct(num) from dis_group | | 7 | 0.00004400 | select num from dis_group group by num | | 8 | 0.00004700 | select distinct(num) from dis_group | | 9 | 0.00004400 | select num from dis_group group by num | | 10 | 1.00069100 | create index num_index on dis_group(num) | | 11 | 0.00362300 | select num from dis_group group by num | | 12 | 0.00200000 | select distinct(num) from dis_group | | 13 | 0.00091000 | select distinct(num) from dis_group | | 14 | 0.00098700 | select num from dis_group group by num | | 15 | 0.00090600 | select num from dis_group group by num 我的结果中，没加索引前速度更快。估计我的机子太烂了~~ 存储过程
DELIMITER $$

DROP PROCEDURE IF EXISTS `test`.`dis`$$

CREATE DEFINER=`root`@`localhost` PROCEDURE `dis`(num int(10))
begin
declare i int(10);
set i = 1;
while i <= num do
insert into dis_group (num) values (floor(rand()*100));
set i=i+1;
end while;
end$$

DELIMITER ;
| Query_ID | Duration | Query |
+----------+------------+------------------------------------------+
| 1 | 0.00099400 | show variables like "%pro%" |
| 2 | 0.09483400 | select distinct(num) from dis_group |
| 3 | 0.09505400 | select num from dis_group group by num |
| 4 | 0.00004600 | select num from dis_group group by num |
| 5 | 0.00004200 | select distinct(num) from dis_group |
| 6 | 0.00004400 | select distinct(num) from dis_group |
| 7 | 0.00004400 | select num from dis_group group by num |
| 8 | 0.00004700 | select distinct(num) from dis_group |
| 9 | 0.00004400 | select num from dis_group group by num |
| 10 | 1.00069100 | create index num_index on dis_group(num) |
| 11 | 0.00362300 | select num from dis_group group by num |
| 12 | 0.00200000 | select distinct(num) from dis_group |
| 13 | 0.00091000 | select distinct(num) from dis_group |
| 14 | 0.00098700 | select num from dis_group group by num |
| 15 | 0.00090600 | select num from dis_group group by num

我的结果中，没加索引前速度更快。估计我的机子太烂了~~

]]>