[f-nsp] running vpls with multiple vlans on CES2000

Lazuardi Nasution mrxlazuardin at gmail.com
Wed Aug 3 18:03:53 EDT 2011


Hi Igor,

Of course like that. On the given script, you connect two VLAN via
VPLS cloud, you can say it VLAN translation. Sometime I do like this
to connect VLANs which come in to different switches/routers.

You should try Q-in-Q  for transporting user VLANs inside VPLS.
External VLAN is provider VLAN (your side), inner VLAN is user VLAN.
Using more than one VPLS cloud can be used but more administrative
overhead since usually single VPLS cloud is related to single
client/project.

Best regards,

> ---------- Forwarded message ----------
> From: Igor Ybema <foundry-nsp at ergens.org>
> To: foundry-nsp at puck.nether.net
> Date: Tue, 2 Aug 2011 20:10:11 +0200
> Subject: [f-nsp] running vpls with multiple vlans on CES2000
> Hi there,
>
> Today I was trying create a VPLS between some CES2000's holding more
> than one vlan so that I could transport two vlan's from the same
> customer over one vpls:
>
> router mpls
>  vpls test 10
>  vpls-peer some-ip-address
>    vlan 100
>      tag e 1/1
>    vlan 200
>     tag e 1/1
>
>
>
> The config would let me do this (instead of with a VLL, which blocks
> when you enter more than one vlan).
> However, when I connected a normal vlan switch on the e1/1, I noticed
> that the mac-addresses from vlan 100 where flooded over vlan 200 and
> vice versa.
>
> Is this indeed not possible? Was I just creating a vlan-translation
> thing? The document is very unclear about this.
> What if I had a MLX/XMR and would do the same with 'vc-mode tagged'
> enabled within the vpls? Would that work?
>
> I fixed my problem by creating two VPLS instances each with one vlan.
>
> regards, Igor Ybema




More information about the foundry-nsp mailing list