> How would I know?  The Solaris system could have been running a few
> thousand processes, handling much more overall load than the 
> Wintel server when your benchmark was run.  What did you do to
> test performance?

Both machines were ours and ours alone for the duration of the first stage of 
testing.

We didn't stop any un-needed tasks on either machine. We ran our application on 
both set-ups and emulated usage for live interaction, programmed interaction 
and recorded interaction.

We then ran the same tests as the second stage of testing on the systems while 
they were being used for their normal day-to-day functions.

Results were in favour of SQLServer in both instances.






Return to Archive home page | Return to MIDRANGE.COM home page