我不是最强的SQL,但我一直在管理.....我有一种情况,其中一个我的内连接返回多行,我试图找到一种方法来抓取行与使用名为"确定性"列的最高值。
语句涉及两个表dim_asset和dim_asset_operating_system。这些表由一个名为asset_id的公共字段和一个名为os_description的字段在dim_asset和一个名为description的字段在dim_asset_operating_system连接。
代码:
SELECT
A.ip_address,
A.os_type,
A.os_vendor,
A.os_family,
A.os_name,
A.os_version,
A.os_description,
A.os_system,
B.certainty
FROM
dim_asset A
INNER JOIN
dim_asset_operating_system B ON A.asset_id = B.asset_id
AND A.os_description = B.description
WHERE
ip_address = '192.168.65.100';
输出:
ip_address | os_type | os_vendor | os_family | os_name | os_version | os_description | os_system | certainty
----------------+---------+-----------+-----------+---------+------------+--------------------+--------------+-----------
192.168.65.100 | | Ubuntu | Linux | Linux | 16.04 | Ubuntu Linux 16.04 | Ubuntu Linux | 1
192.168.65.100 | | Ubuntu | Linux | Linux | 16.04 | Ubuntu Linux 16.04 | Ubuntu Linux | 1
192.168.65.100 | | Ubuntu | Linux | Linux | 16.04 | Ubuntu Linux 16.04 | Ubuntu Linux | 0.85
(3 rows)
理想……我希望返回以下内容:
ip_address | os_type | os_vendor | os_family | os_name | os_version | os_description | os_system | certainty
----------------+---------+-----------+-----------+---------+------------+--------------------+--------------+-----------
192.168.65.100 | | Ubuntu | Linux | Linux | 16.04 | Ubuntu Linux 16.04 | Ubuntu Linux | 1
在上面的示例中,这只是一个记录…理想情况下,我需要将其扩展到返回的数百万行中的10行。
谢谢你提供的任何帮助。
如果您希望获得最高的确定性,您只需要ORDER BY
它并将结果集约束为FETCH FIST ROW ONLY
:
SELECT * FROM dim_asset A
INNER JOIN dim_asset_operating_system B ON A.asset_id = B.asset_id AND
A.os_description = B.description
WHERE ip_address = '192.168.65.100'
ORDER BY certainty DESC
FETCH FIRST ROW ONLY;
然而,如果多个记录共享相同的最高确定性(平局),则此查询将只返回其中一个,这在许多用例中不是理想的行为。如果您对所有共享最高值的记录感兴趣,请使用FETCH FIRST ROW WITH TIES
(在PostgreSQL 13中引入):
SELECT * FROM dim_asset A
INNER JOIN dim_asset_operating_system B ON A.asset_id = B.asset_id AND
A.os_description = B.description
WHERE ip_address = '192.168.65.100'
ORDER BY certainty DESC
FETCH FIRST ROW WITH TIES;
Demo:db<>fiddle