Question

I have a SSIS project that queries a database and get item level data.

The data should be exported to two flat files, one being a summary file,
and the second one a details file. Both files are linked using a index (string)
column that is generated during the SSIS dataflow.

Now i basically want a “distinct index” of the detail data for the summary file.
It is however not a plain aggregate, some columns are strings and instead of a count
I want something like “first(fieldname)” (like in SSRS).

Any suggestions how to do this?

Thanks!

Answer 1

First is just something like ‘Select TOP 1 * from aggregate’.
There is no problem if you can use below query on your source.
If not you need to emulate this somehow in your data flow. (Lookups, Merge Joins etc.)

Here is example in sql:

select a.id, (select top 1 b.name from tableA b where a.id = b.id)
from tableA a
group by id