views:

47

answers:

1

Okay, I am pretty sure at this point that this cannot be done, but here goes:

I have a production SQL Server 2005 database with two filegroups, PRIMARY and SECONDARY. PRIMARY consists of about 2 gigs of typical database data. SECONDARY consists of 15 gigs of archived data that no longer changes.

I would like to create a QA database in such a way that I can use essentially a "static" backup of the SECONDARY filegroup plus a standard full+logs backup of the PRIMARY filegroup to create a working version of production in QA.

All of this is basically to reduce the backups from the full 17 gigs they currently are to a 2 gig filegroup backup which combined with the "static" backup still looks like production.

However, all of my attempts to Google this or attempt it using the database restore methods available in SQL Server 2005 either leaves me in a non-recovered state or gives me grief about LSN discrepancies.

I think I may just be screwing up the syntax with regards to the RESTORE statement, but if it can't be done, I'm willing to accept that, just trying to save me some time and disk space.

A: 

Were you able to get any answer from this scenario you gave?

I am actually having this same scenario but a bigger size of the database, but overall, the idea is the same. Being able to restore the database in QA but not having the same volume of data as it is in Production.

My database has 4 partitions, PRIMARY, FG1, FG2, FG3 (FG2 and FG3 are READ-ONLY), and after one full backup, just doing Partial backups subsequently. When I restore it in QA, I only want PRIMARY and FG1, and the database is already in recovered state. not having FG2 and FG3 would save me more than 150GB of historical data.

I am hoping that it CAN be done. Just needs some testing and clarifications...

Thanks!

sqlN00bie