Alex Rousskov wrote:
> Now I see where the confusion is coming from. You probably
> have a box that is a dedicated load balancer (as opposed to a switch
> that also does load balancing).
yes, however I can come out of the balancer with several interfaces
on each side if I want to, so it will be doing a bit of switching as well.
As to what we call it - dedicated load balancer or whatever, I'm not
In this case, you will most likely
> have to provide a L2 switch(es), at least for the server side, as a
> part of your benchmarking zone. You can use any switch(es) you like.
> We may be able to give you a spare switch from our lab if needed, but
> that will not change the reporting side.
OK. I just need to know how to connect to your setup and where my
zone ends and your zone starts.
I may just connect on both sides (to server, to client)
with multiple crossover cables, don't know yet.
I'll wait till the problem of single -v- multiple cables is
> In the past, the price for that extra equipment was included
> into the total product price, affecting price/performance ratios. In
> your case, the price of a simple L2 switch is unlikely to have a
> serious affect. Also, the official report will note that customers
> that already have necessary gear should adjust price and
> price/performance figures accordingly.
yes, a 100Mbps switch is not a concern for price.
> The question of Gbit versus 100Mbit cables on client and
> server side of the participant zone remains unresolved. IMO, we need
> to try to get participant consensus on this issue. However, whatever
> the final decision is, I hope we can accommodate low-end entries.
> few report readers have 10GBit worth of traffic going through their
> Web server farm...
not happening here :-)
Our goal is to use a realistic setup that can
> accommodate most of the products in a fair fashion.
thanks, looking forward to hearing more as the issue progresses.
-- Joseph Mack PhD, Senior Systems Engineer, Lockheed Martin contractor to the National Environmental Supercomputer Center, mailto:email@example.com ph# 919-541-0007, RTP, NC, USA
This archive was generated by hypermail 2b29 : Tue Jul 10 2001 - 12:00:18 MDT