May 1, 2009 at 10:05 am
We have a large database (17TB divided on 80 filegroups - 11 SAN drives - SQL 2000) that will grow even more. Every month I create a new filegroup (which will be about 300 to 600GB at the end of the month) and the previous one is set to read-only as users don't need to update it.
Since this database generates log backups of 20GB every 1/2 hour, it is almost impossible to keep track of them... we delete them.
As a result, we can't have a reliable backup nor a disaster recovery strategy. Even worse, if one of the disks that hold just one filegroup gets corrupted, the entire database would be offline.
So, we decided to instead of creating new filegroups we will create new databases (and move the old filegroups to databases as well) in this way we can have a manageable backup strategy and we won't take the risk of a partial hardware failure.
At the end we would have the 'original' database (our current primary filegroup which contains the common/core tables and objects) and a bunch of databases in the same server (one per month).
Our concern now is performance:
What would be the performance difference if I query data from a filegroup that is on a different drive of the primary filegroup versus querying another database which filegroup sits on a different drive?.
This is the following step of this topic:
May 1, 2009 at 10:20 am
There is a security check, and you'd need to link in the databases (linked server or view), but there isn't a substantial difference I'd think in having separate databases.
Viewing 2 posts - 1 through 1 (of 1 total)
You must be logged in to reply to this topic. Login to reply