I'm looking for a source of performance information for SQL Server (Express, though I assume it's the same).
I need to know how going from 1,000 x 1000 byte records to 100,000 x 1000 byte records will affect CRUD response times.
I recognize there's always the "it depends"; I'm hoping there are some sources of baseline performance metrics out there that I can extropolate from.
Any guidance on this would be appreciated.
Thanks!
The easiest way to do this is to test it yourself and it really does depend...
You need to test it on the hardware you will be using, disk perf and memory make a HUGE difference.
What indexes do you have? Is the data being inserted going to change the clusterkey? Is there a cluster key? etc etc
Its failr easy to write a little loop on the server or the client that will insert 100,000 rows, try using DTS/SSIS if you want.
|||Belated thanks for the post. Yes, I finally came to understand that is really does depend.
I did an insert and lookup test with nothing else running, and then ran it with four Office Apps loading, and yes, it made a huge difference.
I guess the OPS guys are earning their pay!
sql
No comments:
Post a Comment