mysql查询采用＆gt; 4min进行简单连接

Question

我想知道我能做些什么来加快这个问题。

该机器是带有5900RPM HD的i7。有什么我能做的吗？ collision sha1确实有一个索引，其余的连接在主键上

mysql> select count(*) from (
    -> select f.id, d1.name as dir, n1.name as name, c.sha1, size from file f
    -> join (select sha1 from collision group by sha1 having count(*)>1) msha on 1=1
    -> join collision c on c.id = f.id and c.sha1 = msha.sha1
    -> join NameList n1 on n1.id=f.name
    -> join directory d1 on d1.id=f.dir
    -> ) z;
+----------+
| count(*) |
+----------+
|    66469 |
+----------+
1 row in set (4 min 16.80 sec)

mysql> select count(*) from Collision;
+----------+
| count(*) |
+----------+
|   205713 |
+----------+
1 row in set (0.07 sec)

mysql> select count(*) from (select sha1 from collision group by sha1 having count(*)>1) z;
+----------+
| count(*) |
+----------+
|    29915 |
+----------+
1 row in set (1.98 sec)

mysql> explain select count(*) from (
    -> select f.id, d1.name as dir, n1.name as name, c.sha1, size from file f
    -> join (select sha1 from collision group by sha1 having count(*)>1) msha on 1=1
    -> join collision c on c.id = f.id and c.sha1 = msha.sha1
    -> join NameList n1 on n1.id=f.name
    -> join directory d1 on d1.id=f.dir
    -> ) z;
+----+-------------+------------+--------+------------------+---------+---------+---------------+--------+---------------------------------+
| id | select_type | table      | type   | possible_keys    | key     | key_len | ref           | rows   | Extra                           |
+----+-------------+------------+--------+------------------+---------+---------+---------------+--------+---------------------------------+
|  1 | PRIMARY     | NULL       | NULL   | NULL             | NULL    | NULL    | NULL          |   NULL | Select tables optimized away    |
|  2 | DERIVED     | <derived3> | ALL    | NULL             | NULL    | NULL    | NULL          |  29915 |                                 |
|  2 | DERIVED     | c          | ALL    | PRIMARY          | NULL    | NULL    | NULL          | 265854 | Using where; Using join buffer  |
|  2 | DERIVED     | f          | eq_ref | PRIMARY,dir,name | PRIMARY | 8       | filedb.c.id   |      1 |                                 |
|  2 | DERIVED     | n1         | eq_ref | PRIMARY          | PRIMARY | 8       | filedb.f.name |      1 |                                 |
|  2 | DERIVED     | d1         | eq_ref | PRIMARY          | PRIMARY | 8       | filedb.f.dir  |      1 |                                 |
|  3 | DERIVED     | collision  | ALL    | NULL             | NULL    | NULL    | NULL          | 265854 | Using temporary; Using filesort |
+----+-------------+------------+--------+------------------+---------+---------+---------------+--------+---------------------------------+
7 rows in set (4 min 11.97 sec)