使用union关键字时，可以给出多条select 语句，并将它们的结果合成单个结果集。合并时两个表对应的列数和数据类型必须相同，每个select 语句之间使用union或union all 关键字分隔，union 执行的时候删除重复的记录，所有返回的行都是唯一，使用union all 关键字的作用是不删除重复行也不对结果进行自动排序。
例如：
mysql> select s_id,f_name from fruits where s_id=101 union select s_id,f_name from fruits where f_price<10;
+------+------------+
| s_id | f_name |
+------+------------+
| 101 | apple |
| 101 | blackberry |
| 101 | cherry |
| 103 | apricot |
| 106 | ahfjwj |
| 103 | cococut |
| 102 | grape |
+------+------------+
7 rows in set (0.00 sec)
mysql> select s_id,f_name from fruits where s_id=101 union all select s_id,f_name from fruits where f_price<10;
+------+------------+
| s_id | f_name |
+------+------------+
| 101 | apple |
| 101 | blackberry |
| 101 | cherry |
| 101 | apple |
| 103 | apricot |
| 106 | ahfjwj |
| 101 | cherry |
| 103 | cococut |
| 102 | grape |
+------+------------+
9 rows in set (0.00 sec)
每个select集是这样的
mysql> select s_id,f_name from fruits where s_id=101;
+------+------------+
| s_id | f_name |
+------+------------+
| 101 | apple |
| 101 | blackberry |
| 101 | cherry |
+------+------------+
3 rows in set (0.00 sec)
mysql> select s_id,f_name from fruits where f_price<10
-> ;
+------+---------+
| s_id | f_name |
+------+---------+
| 101 | apple |
| 103 | apricot |
| 106 | ahfjwj |
| 101 | cherry |
| 103 | cococut |
| 102 | grape |
+------+---------+
6 rows in set (0.00 sec)
很明显 union all 没有去重，查询的结果还有重复的行，
