Are you hosting  the server only, or will you also be hosting working
copies? Spread working copies across spindles and put the server on its
own spindle to reduce latency. 

JLM

Reply via email to