RE: [RPRWG] RPR multiring structures
If you do 802.1D/Q bridging on RPR as recommeded by
P802.17, you lose the spatial reuse property of RPR,
regardless of whether it's a single ring or multi-ring
solution. A bridge between RPR rings would operate the
same way as a bridge between Ethernet or token ring
segments.
When you have "too many" nodes on a single ring, you
could potentially run into situations where you have
a congestion point that constantly bounces around
because there are more nodes interacting with one
another. This may be one of the reasons to create
smaller rings. I am not aware of any studies that
recommend a certain maximum value for the number of
nodes on the ring, but that doesn't mean they don't
exist. Quantifying the performance of the fairness
algorithm as the number of nodes on the ring is
increased is probably nteresting exercise.
-Anoop
> -----Original Message-----
> From: Krzysztof Durałek [mailto:kduralek@xxxxxxxxxxxxx]
> Sent: Friday, April 25, 2003 11:53 AM
> To: stds-802-17@xxxxxxxx
> Subject: [RPRWG] RPR multiring structures
>
>
>
> Hello WG members,
>
> I'm a telecommunication student, and I am writing M.A thesis about
> "Interaction between multiring structures in RPR networks". My
> question is: "Is there any advantages or not, about choosing multiring
> net instead single ring?"
> Let imagine specific situation: we have 10-15 nodes, which creates
> campus network - and now I have to find proper solution, then
> next - prove my
> choice: single or multiring.
> I'v got some ideas about this topic, but I couldn't find any results
> of simulating those networks ( multiring ). There are small amount of
> documents describing RPR bridging in general, so I draw a conclusion
> that maybe this way of considering RPR networks is wrong?
> Could you show me right direction ? Maybe some articles about ?
>
> best regards,
> Krzysztof Duralek
> -----------------
> Department of Telecommunication
> AGH - University of Mining and Metallurgy
> Cracow, Poland
>
>