senthil kumar d (7/30/2014)
We have a large Datawarehouse and the size is 50TB.. The tables are placed in filegroups based on the schema like fact, dimensions, raw data each sit on seperate filegroups. I am thinking will it make sense to seperate the large facts which are having billions of rows so that they reside on filegroups on their own.. I would like to know if this is sensible and the pros and cons..
Adding to what has been said and giving the case of how common SANs are now...
if all your mdf files reside on the same LUN (regardless of how many drive letters you may see in the Os), you won't get any I/O improvement anyway, even if you use FGs. This is because in most cases, the whole LUN that contains all the logical drives is already a single tier with same RAID.
By the way, why are you using SIMPLE recovery model? Is your company aware of possible data loss in case of a system crash/restore if you use SIMPLE?