From: Venkataramanaiah.R (vramanaiah@gmail.com)
Date: Fri Jun 02 2006 - 06:11:50 ART
Btw, i can see this dia of 7 is talked about in a MSTP document that
was shared by our fellow groupstudy member. But there are also
instances where the default hop count is 20. I am still researching
what the max can be.., however, i am also thinking MST regions..
What if i split the whole ring into two regions, do you think it will help..
Regarding the MAC address table explosion, it is not a big concern,
because, all these switches are going to be 12 port and i dont see
more than 4-5 hosts connecting to the same switch.
Of course, we end with problems, we can always split the ring with
multiple L3 domains..
-Venkat
On 6/2/06, Venkataramanaiah. R <vramanaiah@gmail.com> wrote:
> Thanks for all your thoughts... and I agree with you all. Actually i
> am working on a proposal, where the customer has recommended a ring
> with 30 switches.
>
> Of course, i am yet to talk to the customer, but before that, i would
> like to make sure, that it is not a good idea to connect so many
> switches in a ring..
>
> Btw, the dia of 7 is probably a limit only for STP. With RSTP, we can
> go for more.. It works well with 12 switches... Want to know, if
> anyone has done more ... :-)
>
> Rgds
> -Venkat
>
>
>
> On 6/2/06, asadovnikov <asadovnikov@comcast.net> wrote:
> > Venkat,
> >
> > Switches in ring topology is bad idea. There following technologies going
> > to work acceptably over such topology:
> > - Sonnet (best option)
> > - Routed network (acceptable option)
> >
> > I much rather do not see rings all together, but if you must due to money
> > limitation do one of the above, do not do switching.
> >
> > I would be surprised if you found case studies... I had seen it done
> > multiple times, and other then saving money it was bad... and nobody wants
> > to put case study out on how bad something works.
> >
> > Best Regards,
> > Alexei
> >
> >
> > -----Original Message-----
> > From: nobody@groupstudy.com [mailto:nobody@groupstudy.com] On Behalf Of
> > Venkataramanaiah.R
> > Sent: Thursday, June 01, 2006 3:01 PM
> > To: James Ventre
> > Cc: ccielab@groupstudy.com
> > Subject: Re: Switches in a ring topology....
> >
> > James, Good sense of humour, and i am also for what you are
> > suggesting, but i think you do not realize that at times due to
> > geographical limiations and to cut cost, people do opt for ring
> > topo... so i am looking for some real life experience or some pointers
> > to some real life case studies :-)
> >
> > Thanks
> > -Venkat
> >
> > On 6/1/06, James Ventre <messageboard@ventrefamily.com> wrote:
> > > Typically, The best practice is to do triangles. You do a triangle with
> > > Your Switch, STP Root and Secondary Root. Rings, Squares, Boxes,
> > > Octagons, etc. generally have higher convergence times when failures
> > > happen and can produce some suboptimal paths depending on how many nodes
> > > are involved.
> > >
> > > James
> > >
> > >
> > >
> > > Venkataramanaiah.R wrote:
> > > > Has anyone implemented such a topology... If so, whatz the max count
> > > > of switches you have
> > > > in the ring..?
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
> >
> > _______________________________________________________________________
> > Subscription information may be found at:
> > http://www.groupstudy.com/list/CCIELab.html
This archive was generated by hypermail 2.1.4 : Sat Jul 01 2006 - 07:57:31 ART