Have you checked the error logs?

Ha! I was just coming down to say something similar. I suspect we're going to be...

Dealing with a super huge table that is partially partitioned can be a challenging task for many database administrators and developers. However, with the right approach and tools, you...

• This reply was modified 3 years, 6 months ago by Jonathan AC Roberts. Reason: link updated

You need to first work out which query is taking the longest then look at optimising that.

You could put some logging in the query to write to a table how...

**************************************** cross tab

Table '#TestTable'. Scan count 5, logical reads 40651, physical reads 0, page server reads 0, read-ahead reads 0, page server read-ahead reads 0,...

Thank you all!

Sorry but I made a mistake on my original posting. Revised DATA representation is below

IF OBJECT_ID('tempdb..#TestTable') IS NOT NULL
  ...

Set up some data (you should have done this when asking the question)

DROP TABLE IF EXISTS #temp
GO
SELECT *
  INTO #temp
  FROM (VALUES (1, 'ON_001', '12/5/2022'),
...

Here is my attempt:

That worked and is reasonably efficient.

You can just schedule the job to run every 5 minutes starting at 7 am until a time when you say the file has failed to arrive.

Freakin' awesome test post, Jonathan.  Well done, good Sir!!!  This also proves something else... code written to do a specific task is highly likely to blow generic methods out...

I think you can beat any splitting function by splitting in code with no need for aggregation.

As there are only a maximum of 4 values you...

You could consider using DBCC CHECKDB with the "PHYSICAL_ONLY" option, which only performs a physical consistency check and not a logical consistency check. This can significantly reduce the memory requirements...

I think you can beat any splitting function by splitting in code with no need for aggregation.

As there are only a maximum of 4 values you can safely...

Yes... and they're making an big deal about it.  ChatGPT is constantly providing incorrect answers about code...

I think you can beat any splitting function by splitting in code with no need for aggregation.

As there are only a maximum of 4 values you can safely use this...