T-SQL选择具有相同列值的完整行

Question

I want to show all rows that has a serial that exist in another row.

If I do like this it works

SELECT 
      [Serial]
FROM [x].[dbo].[Devices]
GROUP BY Serial
HAVING COUNT(*) > 1

But when I add more select columns

SELECT [ID]
      ,[UUID]
      ,[Serial]
FROM [x].[dbo].[Devices]
GROUP BY Serial
HAVING COUNT(*) > 1

I get

'ID' is invalid in the select list because it is not contained in either an aggregate function or the GROUP BY clause.

Why can't I select more columns?

How am I suppose to show the full rows?

Answer 1

You can do this with a window function:

select *
from (
   select *, 
          count(*) over (partition by [Serial]) as serial_count
   from [x].[dbo].[Devices]
)  t
where serial_count > 1;

This is typically faster then joining to a sub-select with an aggregate.

Answer 2

Hope you need this query, it will show all the rows that has a serial that exist in another row.

SELECT  D1.[ID]
        ,D1.[UUID]
        ,D1.[Serial]
FROM [x].[dbo].[Devices] D1
JOIN (  SELECT [Serial]
        FROM [x].[dbo].[Devices]
        GROUP BY Serial
        HAVING COUNT(*) > 1 ) D2 ON D1.[Serial] = D2.[Serial]

Answer 3

As long as ID and UUID are unique to the serial, try grouping by all columns.

SELECT [ID]
  ,[UUID]
  ,[Serial]
FROM [x].[dbo].[Devices]
GROUP BY Serial
  ,[ID]
  ,[UUID]
HAVING COUNT(*) > 1