Re: VPLS Split Horizon Doubt

Are you trying to understand why it is when a BUM (broadcast,
unknown-unicast and multicast) frame is received on a pseudowire that is in
a mesh-group, it will not send the frame out to other mesh-group members?
Primarily this is to help contain replication through the VPLS, otherwise
you would achieve a nice self induced denial of service attack.

The split horizon rule for BUM frames in VPLS just means you treat all the
psuedowire of the mesh-group as if they are a single big psuedowire. The
mesh group is configured as a full mesh, so if something came in from a
mesh-group psuedowire, all the other nodes that are part of the mesh-group
would have recieved that frame too.

The two types of pseudowires in a VPLS that I'm aware of are spoke (just
your regular point-to-point) and mesh. You can combine these in interesting
ways to build a hierarchical VPLS (H-VPLS) which means you can have large
topologies but not have a N squared problem when it comes to putting
together the psuedowires to build it.

Cheers,
Adam

On Wed, Dec 7, 2011 at 4:30 PM, Routing Freak <routingfreak_at_gmail.com>wrote:

> Hi all,
>
> I have a doubt in VPLS Split Horizon rule. My understanding is " Whenever a
> frame received from a psuedowire , that particular frame should not be sent
> o anohter Psuedowire . So inside the same Bridging domain,t here can be
> many psuedowires, so how could a loop form when there are mulitple
> pseudowires inside the same bridged domain?. So how many types of
> pseudowires are present in VPLS?
>
> Please correct me if i am wrong.
>
>
> 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 Dec 07 2011 - 19:43:12 ART