Server Fault is a question and answer site for system and network administrators. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am currently working on a uni project about routing protocols and network performance, one of the criteria i was going to test under was to see what effect lan topology has, ie workstations arranged in mesh, star, ring etc, but i am having doubts as to whether that would have any affect on the routing performance thus would be useless to do, rather i'm thinking it would be better to test under the topology of the routers themselves, ie routers arranged in either star, mesh ring etc.

I would appreciate some feedback on this as I am rather confused.

Thank You

share|improve this question

I think you need to look a little closer at what performance you are trying to measure. Are you trying to measure the performance of routing protocol convergence time, network throughput or something else?

I suppose one project would be looking at the time it may take for a router to choose another path when the current path goes down. You could look at the route update differences betwen various routing protocols in different topology designs? However, I suppose this sort of thing is already well documented for various routing algorithms/implementations.

I don't think a router will care too much what sort of LAN it is part of... star, ring, mesh. It will build its forwarding table according to what it hears from other routers on the network and will choose a best path to the other networks.

You say you're a bit confused so it maybe best to work a little on the question and what you want to test before jumping into it.

share|improve this answer
Thank you for your reply In regards to network topology scenarios i will be looking at throughput and traffic sent received (bps), packet delay, basically the opnet software provides options for statistics and these are some of them, i will do separate tests on convergence put not in these scenarios. However what you wrote seems like a good suggestion so i might take that onboard. I am quite confused lol, the criteria and aims i set up for the project seem complex now and i am sort of learning as i go along, hence the question i put up as i am unsure whether it will be worth doing that. – algorithms Apr 2 '10 at 22:09

What does your test bed look like? With todays bandwidth you are going to need a lot of PC's / LAN analyzers to put enough load on the network to notice.

Also, when you say performance do you include redundancy part of that, or are you just looking at raw throughput?

share|improve this answer
thank you for your reply I will be testing using a simulation software by opent, to give one example i will be testing using 50 users/workstation under one scenario, the lan will be segmented into subnets, and the topology of the lan could be say mesh or star, and all the lans will be connected using routers, connection bandwidth will be 100bt for one scenario, but what i want to know is would the network topology of a lan have any bearing on the performance of routing in the routers, the criteria i will look at are quite a few including throughput, packets sent and received etc – algorithms Apr 2 '10 at 18:21

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.