RE: Discussion on E-Tree and H-VPLS

Sasha,

Just as we discussed for many times, 3 approaches use different identifiers.
For me, Multi-PW uses different PATH (Multi-PW), but CW or Dual-VLAN uses
inferring context in frames.

So CW also has same HVPLS deployment problem.

Thanks,

Sam

-----Original Message-----
From: Alexander Vainshtein [mailto:Alexander.Vainshtein@...]
Sent: Monday, April 23, 2012 7:33 PM
To: Jiangyuanlong; Sam Cao
Cc: l2vpn@...
Subject: RE: Discussion on E-Tree and H-VPLS

Sam, Yuanlong and all,
From my point of view the CW-based solution is by far the simplest way to
deal with H-VPLS: all that is required is preservation of the CW (or, in the
unlikely case PW sequencing is used, of its flags) when a packet is
forwarded from the "core" PW to a "spoke" one (or vice versa).

With Dual PWE the situation becomes more complicated IMO. Not sure about
dual VLAN - but may be also tricky.

Regards,
     Sasha
topic.
>
> Regards
> Yuanlong
>