[c-nsp] VPLS, spanning tree and redundancy

[Mateusz Błaszczyk]

Pshem,

>  > Can I ask why 2 VPLSes instead of one? Is there any technical reasons ?
>
>
> The main one is so BPDUs from CSW3 and CSW4 can't reach CSW1 at the
>  same over the same link - a single VPLS cloud would act like a hub
>  without spanning tree, forwarding packets in every direction that
>  effectively would stop the L2 network.

So if you do a split like this why using VPLS? why not EoMPLS xconnects (2x) ?


Best Regards,

-- 
-mat