更好地查询“选择”中的“选择”

Question

是否有更好和更短的方法来产生如下所示的SQL查询？较短的查询或/和影响最快的性能发送到服务器的效果更好

select TowerID, COUNT(TowerID) as 'QTY'
from (
    select TowerID, [USER] as 'QTY'
    from ColoEngDBLocal..Carriers
    where [Status] not like '%removed%'
    group by TowerID, [USER]
) as a
group by TowerID
order by TowerID

当我在小数据上使用此查询时似乎没有什么区别，但是当我在大数据上使用此查询时，查询变慢。 （我的行超过60万）

目前，我有如下所示的表

注意：我正在使用SQL Server 2008 R2

Answer 1

尝试一下：

简单COUNT（DISTINCT ....）将为您提供帮助。

select TowerID, COUNT(DISTINCT [USER]) as 'QTY'
from ColoEngDBLocal..Carriers
where [Status] not like '%removed%'
group by TowerID
ORDER BY TowerId