Question

SELECT MIN( a.severity ), CONCAT_WS(",", a.dev_id, a.object_id ) AS tuple 
FROM net.alerts AS a INNER JOIN net.deviceinfo AS c ON c.id = a.dev_id AND 
c.peer = ( SELECT value FROM local.settings WHERE setting="server_id" ) WHERE 1 
AND (a.dev_id, a.object_id) IN ( (13,4164),(45,-1),(145,995),(188,-1) ) GROUP BY tuple


SELECT MIN( a.severity ), CONCAT_WS(",", b.device_id, b.object_id ) AS tuple 
FROM net.alerts AS a INNER JOIN local.conditions AS b ON a.condition_id = b.condition_id WHERE 1 
AND(b.device_id, b.object_id) IN ((13,4164),(45,-1),(145,995),(188,-1) ) Group BY tuple

我有这些表，在警报和条件表中，它们都有device_id和object_id，但它们可能不同，所以我试图检索该组合的所有数据。

Answer 1

我不确定我理解你尝试做什么，但我想你需要使用UNION。

顺便说一下，为什么要写WHERE 1 AND condition而不是WHERE condition？

Answer 2

可能最好只处理一个查询，但是UNION会为你工作，因为你说你只想合并结果：

(SELECT MIN( a.severity ), CONCAT_WS(",", a.dev_id, a.object_id ) AS tuple 
FROM net.alerts AS a INNER JOIN net.deviceinfo AS c ON c.id = a.dev_id AND 
c.peer = ( SELECT value FROM local.settings WHERE setting="server_id" ) WHERE 1 AND 
(a.dev_id, a.object_id) IN ( (13,4164),(45,-1),(145,995),(188,-1) ))

UNION 

(SELECT MIN( a.severity ), CONCAT_WS(",", b.device_id, b.object_id ) AS tuple 
FROM net.alerts AS a INNER JOIN local.conditions AS b ON 
a.condition_id = b.condition_id WHERE 1 AND(b.device_id, b.object_id) IN ((13,4164),
(45,-1),(145,995),(188,-1) ))

使用“UNION ALL”允许重复值

Answer 3

您需要使用union或union all，具体取决于您想要的加入类型。因此，对于twi qyery结果n1和n2：

联盟 - x < n1+n2 - 将删除重复项
全部联盟 - x == n1+n2 - 将保存重复项

所以你的代码应该是这样的：

(SELECT MIN( a.severity ), CONCAT_WS(",", a.dev_id, a.object_id ) AS tuple 
FROM net.alerts AS a INNER JOIN net.deviceinfo AS c ON c.id = a.dev_id AND 
c.peer = ( SELECT value FROM local.settings WHERE setting="server_id" ) WHERE 1 
AND (a.dev_id, a.object_id) IN ( (13,4164),(45,-1),(145,995),(188,-1) ) GROUP BY tuple)

 UNION

(SELECT MIN( a.severity ), CONCAT_WS(",", b.device_id, b.object_id ) AS tuple 
FROM net.alerts AS a INNER JOIN local.conditions AS b ON a.condition_id = b.condition_id WHERE 1 
AND(b.device_id, b.object_id) IN ((13,4164),(45,-1),(145,995),(188,-1) ) Group BY tuple)

您需要记住：查询列的数量和名称必须完全匹配。

如何将这两个查询合并在一起？

3 个答案: