I have solved the problem by adding an IGNORE_DUP_KEY UNIQUE index to my primary key field. This means that I can use Fast Load, without having to check the field first!
However, any duplicates then set the TDT to a failed status, something that I don't want. How can I check for a specific...
I'm running SUSE 9.1 Personal that I burnt from the ISO available from their homepage. That was after trying Xandros (hmmmm...) and PCLinuxOS. PCLOS is good for trying out Linux, as it doesn't install and can run from a CD. If you like it, you can then install it to your HD.
The code is :
SELECT field1
FROM [Table Name]
WHERE tableid = ?
I then pass the tableid into the lookup to see whether it returns the field I chose. Apologies for putting dummy names, but some of the structure could be company-sensitive!!
I had thought of that, but the problem I have is that this task has been written so that any number of them can be run in parallel.
That would mean using my unique process identifier to mark the incoming records in the staging table and then do a distinct select, all of which seems like a lot...
I have a Transform Data Task that copies a lot of data from my source system. Unfortunately, I cannot use a DISTINCT in the SQL from the source system, due to a very poor ODBC driver!
So, when I am creating my primary key, I am trying to do a lookup on the PK column before I insert the record...
This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
By continuing to use this site, you are consenting to our use of cookies.