[c-nsp] Q. About spanning tree (pvst) behavior on hybrid cat6500

Adrian Pirciu adrian.pirciu at rdsnet.ro
Tue Nov 9 04:20:33 EST 2004


On Monday 08 November 2004 20:29, Matt Stockdale wrote:
> Hello -
>
>   Just a brief question about how spanning tree (pvst) behaves on a
> hybrid cat6500 (running catos 7.6.9 for argument).
>
>  Let's say that
>
> a) this 6500 is a core switch, with many (50+) 802.1q trunks to
> access switches (2950's)
> b) There is another core 6500, to which all vlan's are trunked,
> except vlan 100.
>
> My question is this. If the MSFC2 in said 6500 has a vlan100
> interface configured, and a server connected directly to the 6500 in
> vlan100, Will it retain connectivity to the server (and vice versa)
> in the event that the second core switch fails, causing the 6500 in
> question to reconverge all of the spanning tree instances?

I'm not sure I understood corectly the question, but I'll try an answer:

If the second 6500 fails, not all the spanning trees will reconverge. 
For example, since vlan100 is not trunked to the second 6500, nothing 
in this vlan will change (it's important to understand that pvst acts 
per vlan, not per link, i.e. on a trunk, even if all vlans but 100 are 
blocking or whatever, vlan 100 may be forwarding). The answer is that 
the server should not lose connectivity to first 6500. 

>
> Second, does the answer change if instead of being directly connected
> to the 6500, it's connected to a 2950 that is then connected to the
> 6500? (not with a trunk, just running as a "unmanaged" switch, with
> it's "vlan1" uplink into a port on the 6500 configured as vlan100).

Nothing is changing, too. 

>
> And finally, what if vlan100 _is_ trunked to the other core switch?
> Will the 6500 be smart enough to unblock vlan100 on the virtual trunk
> to it's own MSFC while it may be blocking it on any other trunks?

This i don't understand (especially the virtual trunk thing). Please add 
more details to the description. 

>
> Thanks in advance for your help, or pointers to the documentation I
> overlooked which clearly answers this :)
>
> Matt
>
>
> _______________________________________________
> cisco-nsp mailing list  cisco-nsp at puck.nether.net
> https://puck.nether.net/mailman/listinfo/cisco-nsp
> archive at http://puck.nether.net/pipermail/cisco-nsp/

-- 
------------------------------------------------------------------------
Adrian Pirciu
Network Engineer


More information about the cisco-nsp mailing list