QOD 03/12
-
Good question IMHO, except I'd be a bit concerned if I were the "developer" having to make these decsision...
Of course I'd have a different opinion if I'd got it wrong...
SJT
SJT -
So riddle this to me then, Batman. The optimum query plan would be to perform index seeks across the clustered index on one table, and a non-clustered index on the heavily hit table. Why would we want to put both indexes on the same disk drive? There would be no parallel I/O operations. That stream is what's joining to the clustered index on the heavily used table - not the non-clustered index directly
-
I don't really agree with the answer to the QOD. Microsoft doesn't agree either. Here's what BOL has to say about this (the topic is Placing Indexes on Filegroups)
"Because you cannot predict what type of access will take place and when it will take place, it could be a safer decision to spread your tables and indexes across all filegroups. This would guarantee that all disks are being accessed since all data and indexes are spread evenly across all disks, no matter which way the data is accessed. This is also a simpler approach for system administrators."
Just my $0.02
-
quote:
Because you cannot predictOne minor point of distinction. The QOD inferred that you KNOW which tables fell under the category of heavy use and/or multi join. The BOL infers that you do not have this information. In the case of our databases, I know exactly which tables fall in these categories. So the BOL answer that anote2chris suggested does not fully consider all the variables that the QOD addressed. If you did not know the heavy use /multi join tables then the BOL answer would be appropriate.
Steve
Edited by - kepr00 on 12/03/2003 1:04:28 PM
Steve
Viewing 4 posts - 1 through 3 (of 3 total)
You must be logged in to reply to this topic. Login to reply