Network design is an iterative process. I agree with you that it's not entirely linear.
Top-Down Network Design doesn't deal so much with stacks or layers as it does with the need to do a logical design before a physical design. Also, it makes the important point that the first steps must involve identifying the customer's needs and requirements. Sounds obvious, but that's the step that is often overlooked. A lot of engineers think of network design as a Lego project or connect-the-dots game. Many of the software packages that help one do design propagate that thinking. Identifying the customers needs and goals involves many business issues, as you mentioned, but it's also a very technical process. A good network design is based on analysis of existing and future network traffic flow, volume, characteristics, (broadcast, client/server, distributed, errors, windowing), etc. I don't want to rewrite the book in this setting, especially since it's Saturday and the sun is fading, but if we are going to discuss this on a serious level, I may write more later..... Priscilla At 01:42 PM 5/18/02, Kevin Cullimore wrote: >Is it worthwhile to re-examine the assumption that an individual should >start at one end of the all-important "stack" and work their way to the >other end? > >As far as the process itself goes (although, all too often there's not a lot >of process or method associated with design, but I'm sure we're all at least >implicitly aware of that), I'm not sure that the undertakings on behalf of >one layer can be as precisely isolated from one another as the data >structures themselves. If the strategy adopted is bottom-up (thereby >emulating many career paths), I'm not sure that one can tackle layer one >decisions without an understanding of they type & volume of traffic they >will be forced to contend with. For those of you who acknowledge data >connectivity layers > 7, the unreasonable expectations & contradictory >demands of the financial backers would probably need to be addressed first. > >Anyway, I'm just wondering if there exist advantages to working out of >order; I'm well aware that certain advantages exist to working IN order . . >FAQ, list archives, and subscription info: >http://www.groupstudy.com/list/cisco.html >Report misconduct and Nondisclosure violations to [EMAIL PROTECTED] ________________________ Priscilla Oppenheimer http://www.priscilla.com Message Posted at: http://www.groupstudy.com/form/read.php?f=7&i=44452&t=44417 -------------------------------------------------- FAQ, list archives, and subscription info: http://www.groupstudy.com/list/cisco.html Report misconduct and Nondisclosure violations to [EMAIL PROTECTED]
