Great question! :) Rich wrote: > > In short to do a proper benchmark, there needs to be some measure of > comparing apples to apples and ec2 to hardware boxes > is like apples to ducks - not even the same type of creature - IMHO Looking at benchmarks based on efficiency measures of energy consumed and physical space consumed per unit (query, bandwidth, etc.) might give a good way to benchmark the two different hardware schemes. This might be off topic though. Cheers, Charles Uchu Strader > > In my quick opinion anyway, > Rich > > Michael Giles wrote: >> Andy, >> >> #3 seems like a perfect use of Amazon's EC2 service >> (http://www.amazon.com/gp/browse.html?node=201590011). Essentially, >> you get as many dedicated servers as you want and you only pay for >> the time that you use them (10 cents/hour). If you use a server for >> a whole month, I believe it works out to about $70. But if you only >> use it for an hour, or a day, you just pay for that fraction. It is >> great for just this type of work (where you want a relatively large >> number of machines for a short amount of time). >> >> -Mike >> >> Andy Klapper wrote: >>> I am looking for help finding: >>> >>> 1) A MySQL cluster guru that I can talk to for an hour or two about >>> MySQL >>> clustering configurations for performance, and redundancy. Somebody >>> with >>> actual experience with at least five clustered machines only. >>> >>> 2) Recommendations for vendors selling low cost 1 or 2 U Linux >>> servers. >>> >>> 3) I am looking for a low cost way of using 16 to 24 Linux servers >>> using >>> MySQL clustering and custom software for a performance >>> test/verification of >>> concept test. I doubt that this exists, but I'd like to rent the >>> machines >>> for as little as two weeks to a month. >>> >>> >>> Thanks in advance for any help with any of this questions! >>> >>> >>> Andy.