如何在 MySQL 中有效地选择随机记录?
mysql> EXPLAIN SELECT * FROM urls ORDER BY RAND() LIMIT 1;
+----+-------------+-------+------+---------------+------+---------+------+-------+---------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+-------+---------------------------------+
| 1 | SIMPLE | urls | ALL | NULL | NULL | NULL | NULL | 62228 | Using temporary; Using filesort |
+----+-------------+-------+------+---------------+------+---------+------+-------+---------------------------------+
以上并不符合有效的条件,我应该如何正确地做到这一点?
更新
似乎使用答案中提到的解决方案仍然没有帮助:
mysql> explain SELECT *
-> FROM (
-> SELECT @cnt := COUNT(*) + 1,
-> @lim := 10
-> FROM urls
-> ) vars
-> STRAIGHT_JOIN
-> (
-> SELECT r.*,
-> @lim := @lim - 1
-> FROM urls r
-> WHERE (@cnt := @cnt - 1)
-> AND RAND(20090301) < @lim / @cnt
-> ) i;
+----+-------------+------------+--------+---------------+------+---------+------+-------+------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+------------+--------+---------------+------+---------+------+-------+------------------------------+
| 1 | PRIMARY | <derived2> | system | NULL | NULL | NULL | NULL | 1 | |
| 1 | PRIMARY | <derived3> | ALL | NULL | NULL | NULL | NULL | 10 | |
| 3 | DERIVED | r | ALL | NULL | NULL | NULL | NULL | 62228 | Using where |
| 2 | DERIVED | NULL | NULL | NULL | NULL | NULL | NULL | NULL | Select tables optimized away |
+----+-------------+------------+--------+---------------+------+---------+------+-------+------------------------------+
mysql> EXPLAIN SELECT * FROM urls ORDER BY RAND() LIMIT 1;
+----+-------------+-------+------+---------------+------+---------+------+-------+---------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+-------+------+---------------+------+---------+------+-------+---------------------------------+
| 1 | SIMPLE | urls | ALL | NULL | NULL | NULL | NULL | 62228 | Using temporary; Using filesort |
+----+-------------+-------+------+---------------+------+---------+------+-------+---------------------------------+
The above doesn't qualify as efficient,how should I do it properly?
UPDATE
Seems using the solution mentioned in the answer still doesn't help:
mysql> explain SELECT *
-> FROM (
-> SELECT @cnt := COUNT(*) + 1,
-> @lim := 10
-> FROM urls
-> ) vars
-> STRAIGHT_JOIN
-> (
-> SELECT r.*,
-> @lim := @lim - 1
-> FROM urls r
-> WHERE (@cnt := @cnt - 1)
-> AND RAND(20090301) < @lim / @cnt
-> ) i;
+----+-------------+------------+--------+---------------+------+---------+------+-------+------------------------------+
| id | select_type | table | type | possible_keys | key | key_len | ref | rows | Extra |
+----+-------------+------------+--------+---------------+------+---------+------+-------+------------------------------+
| 1 | PRIMARY | <derived2> | system | NULL | NULL | NULL | NULL | 1 | |
| 1 | PRIMARY | <derived3> | ALL | NULL | NULL | NULL | NULL | 10 | |
| 3 | DERIVED | r | ALL | NULL | NULL | NULL | NULL | 62228 | Using where |
| 2 | DERIVED | NULL | NULL | NULL | NULL | NULL | NULL | NULL | Select tables optimized away |
+----+-------------+------------+--------+---------------+------+---------+------+-------+------------------------------+
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。
绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论
评论(2)
Quassnoi 写了一个发布关于随机选择行而不执行排序的内容。他的示例随机选择 10 行,但您可以对其进行调整以仅选择一行。
如果您希望它真正快,那么您可以使用不完全均匀或有时无法返回行的近似值。
您还可以使用存储过程从 Bill Karwin 的帖子:
请注意,这在 MyISAM 中比 InnoDB 运行得快得多,因为 COUNT(*) 在 InnoDB 中很昂贵,但在 MyISAM 中几乎是即时的。
Quassnoi has written a post about selecting rows at random without performing a sort. His example selects 10 rows at random, but you can adapt it to select just one row.
If you want it to be really fast then you can use an approximation that won't be completely uniform or will sometimes fail to return a row.
You can also use a stored procedure to select a random row quickly from Bill Karwin's post:
Note that this will run much faster in MyISAM than InnoDB because COUNT(*) is expensive in InnoDB but nearly instant in MyISAM.
好吧,如果您可以将一些逻辑移至应用程序层(并且我没有误解您的问题),那么您所需要做的就是在应用程序中生成随机 ID,然后对由该键标识的一条记录执行简单的选择。您需要知道的只是记录数。哦,如果该密钥被删除,请获取下一个。
Well, If you can move some logic to application layer (and I didn't misunderstood your question), then all you need is to generate random ID in your application and then perform simple select for one record identified by that key. All you need to know is count of records. Oh, and if that key was deleted, get next one.