Re: interface Vlan vs L3 port switch

From: Peter Alston <palston_at_gmail.com>
Date: Wed, 18 May 2011 22:24:51 +0100

The old KISS adage springs to mind.

Keep It Simple Stupid - why add an extra layer of complexity to a situation
unless absolutely necessary?

just my 2c

P

On Wed, May 18, 2011 at 10:08 PM, ALL From_NJ <all.from.nj_at_gmail.com> wrote:

> Interesting questions.
>
> No STP for L3 ports ... STP is for the L2 links.
>
> Not sure I see the point of having a trunk link between two switches as
> well
> as a L3 port on each connecting each to each other. Maybe there are other
> devices on the network which are broadcast based and cannot be routed?
> Routed traffic across one link and broadcast on the other? Perhaps this is
> to extend a particular vlan?
>
> Maybe this design will allow for rspan across the switches for management
> purposes? Separate L3 and L2-only networks ... Humm ...
>
> Not sure I know why, but I think your questions are good amigo. I would
> keep asking these questions to your colleague and even propose a lab test.
>
> If you learn more of why this design is proposed, this may help to
> understand if it is a sound design.
>
> I do not see any gotchas ... but perhaps the team can think of some. As
> far
> as other concerns and chances for error, I would ask you to consider the
> "usual safe guards" ... disable negotiations, bpduguard, root, port
> security, etc ... etc ..
>
> HTH,
>
> Andrew
>
>
>
>
>
> On Wed, May 18, 2011 at 10:50 AM, Juan <fferrer10_at_gmail.com> wrote:
>
> > Thanks Andrew,
> >
> > The topology is something like this:
> >
> >
> > SW1
> > SW2
> >
> > L3port*SW1* L2port*SW1*(Vlan2) <------Dot1q trunk------>
> L2port*SW2*(Vlan2)
> > L3port*SW2*
> > |
> > |
> > | |
> >
> > |_________|
> > |_____________|
> > Physical
> > cable1 Physical
> > cable2
> >
> >
> >
> >
> > , and obviously the Layer3 ports of the 2 switches are on the same ip
> > network. As you are directly interconecting 2 ports of each switch,
> > is there a possible risk of having a loop, maybe at booting, when the
> > switches are turnning on? And what happens with the STP, may it block
> some
> > port or cause the systems to continously complain about vlan mismatch?
> >
> > I never used a scenario like this, nor in lab environments... to me it
> > seems a "bad" solution, but it would be helpfull if you can point the
> > problems this topology could cause, if you clearly see anyone.
> >
> > Regards
> >
> > Juan
> > 2011/5/18 ALL From_NJ <all.from.nj_at_gmail.com>
> >
> >> Do I understand the topology like this:
> >>
> >> L3port-SW1 <-> L2-SW2 <-> L3port-SW3
> >>
> >> Having any L3 device connect to your L2 switch is fine.
> >>
> >> Having two L3 devices connecting to your L2 switch and configuring these
> >> ports within the same VLAN means that they can talk just fine ;-)
> >>
> >> I would suggest to use portfast on the interfaces connecting to the L3
> >> devices. You may also consider bpdufilter as well as a default for when
> >> using portfast.
> >>
> >> Is it a good design in your network? Humm ... you would know best.
> >>
> >> HTH,
> >>
> >> Andrew
> >>
> >> .
> >>
> >> On Wed, May 18, 2011 at 2:42 AM, <fferrer10_at_gmail.com> wrote:
> >>
> >>> Hi Guys,
> >>>
> >>> I have some doubts about a config a colleague of me is doing using L3
> >>> Catalyst
> >>> switches: in order to have IP communication on 2 linked switches, he is
> >>> using a
> >>> L3 port on each one, and conecting this port (with a cable) to a L2
> port
> >>> of a
> >>> common vlan (in each switch)... As the vlan crosses the switches and
> has
> >>> L2
> >>> ports on each one, he can connect another L3 port, with the same ip
> >>> network, to
> >>> a port of the second switch and have the 2 switches speaking IP in
> this
> >>> vlan
> >>>
> >>> My first impression was "this can be or could be a potential loop, why
> >>> not
> >>> using the Vlan interface, that is done for this cases?"... Anyway,
> since
> >>> he
> >>> told me that is plannig to run BGP and he does not like a vlan
> interface
> >>> as
> >>> bgp speaker, i am wondering if there are reasons to totally reject this
> >>> strange/curious design.
> >>>
> >>> Anyone can provide help to see the possible problems of this scenario
> on
> >>> a
> >>> production environment, please?
> >>>
> >>> TIA and best regards.
> >>>
> >>>
> >>> Blogs and organic groups at http://www.ccie.net
> >>>
> >>> _______________________________________________________________________
> >>> Subscription information may be found at:
> >>> http://www.groupstudy.com/list/CCIELab.html
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>>
> >>
> >>
> >> --
> >> Andrew Lee Lissitz
> >> all.from.nj_at_gmail.com
> >>
> >
> >
> >
> > --
> >
>
>
>
> --
> Andrew Lee Lissitz
> all.from.nj_at_gmail.com
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html

Blogs and organic groups at http://www.ccie.net
Received on Wed May 18 2011 - 22:24:51 ART

This archive was generated by hypermail 2.2.0 : Wed Jun 01 2011 - 09:01:11 ART