SQLIO and SQLIOsim

  • Hi all

    We have a new SAN and I have decided it would be good to try and stress test this to get some numbers. Unfortunately this is a new area for me. I have been looking at Brent Ozars blog on SQLIO and found it a good tutorial on running SQLIO to produce some numbers about how the disk subsystem can perform. This alone raises some questions for me however on the parameters being used and when the correct parameters to use are 8k or 64k blocks and when to use random over sequential.

    For example T-log write sequentially I believe, and OLTP systems write in a more random manor. But what about reads? Can OLTP write in sequential manor? Does this depend on the type of process being executed and what are examples of these? How can I tell if a process is sequential or random?

    The other problem I have is with SQLIOsim. There seems to be very little out there on how to use the tool or how to configure it to replicate SQL Server performance.

    My ideas on all this are all over the place at the moment and I was hoping maybe someone could point me in the right direction.

    Thanks for any input

Viewing post 1 (of 1 total)

You must be logged in to reply to this topic. Login to reply