From: Ken Diliberto (ken@kdmd.net)
Date: Thu Jun 01 2006 - 18:30:19 ART
Venkat,
If you want to stay within the STP specification, the answer is 7. Can
you do more? Sure. Is it safe? No. Have a look at this document (it's
for the 2950, but spanning-tree is spanning-tree):
http://www.cisco.com/univercd/cc/td/doc/product/lan/cat2950/1219ea1/scg/swmstp.htm#1038772
Notice the net-diameter parameter in step 2. This is the *MAXIMUM*
number of switches between two end stations (workstations, servers,
etc). Draw out your topology and count the longest layer 2 distance
between two end stations without duplicating a path. No short-cuts
allowed -- you must measure the longest possible path. If the longest
path is greater than 7 switches, you're over the specification.
What can happen when you're over the specification? Maybe nothing -- if
you're lucky. When talking about real-world networks with real-world
traffic, you could experience random network outages while spanning-tree
reconverges. This could take 30-60 seconds (maybe more) depending on
the network.
To get around this, add layer 3 boundaries and/or design a star topology.
Ken
Venkataramanaiah.R wrote:
> Hi all,
>
> I would like to know, what is the maximum number for 3560 switches
> that can be connected together using a ring topology(With RSTP)?
>
> I know of a setup, which works well with about 12 switches in the
> ring, but now i have
> a customer, who wants to connect about 30 switches in one ring... Is
> this possible? If not, can i break them as two rings? Will it work
> this way?
>
> Has anyone implemented such a topology... If so, whatz the max count
> of switches you have in the ring..?
>
> Please share your experience..!
>
> Regards
> -Venkat
This archive was generated by hypermail 2.1.4 : Sat Jul 01 2006 - 07:57:31 ART