On 10 Oct 2002 15:46:25 -0700 Jack Coates <[EMAIL PROTECTED]> wrote:
> Talk about on-topic -- I'm working with bing right now. Unfortunately, > it's giving some very odd results on a DS-3 line, and they're the kind > on non-deterministic odd that makes one think better numbers could be > procured with a hat and some slips of paper... > > Works fine on DSL lines and cable modems though. Has anyone used this > successfully on high capacity links? > > Please respond off-list and I'll summarize. <flame> Folks, Let's look at the problem in terms of cars and roads... DSL/Cable: like driving up/down your own driveway... easy to determine how may cars you can put on it. DS-3: like trying to determine the traffic capacity of a freeway by sending more cars onto it without knowing the current traffic situation. If you think you can determine bandwidth on a link which is carrying traffic from hundreds/thousands of other users by injecting yet more traffic, you're just not aware of traffic, patterns, shaping, queueing, prioritication, transit times, retransmission impacts, etc, etc, of transmission systems... Adding more load on the link in the form of [pb]ings just further congests the link with USELESS traffic... Worse, the more useless traffic is injected into a network, the more legitimate applications have to retransmit, further wasting bandwith, until the network "avalanches"... "Tools" like this are unscientific and in an indirect way a DDoS attack on the network. Please refrain from using such trash just cuz you want to see a "cool result"... <SIGH> </flame> Pierre
Want to buy your Pack or Services from MandrakeSoft? Go to http://www.mandrakestore.com