Re: Consensus call on adopting <draft-krishnan-6man-rs-mark-08.txt>

[Ole Troan <[e-mail]>]

David,

good point indeed.

perhaps it is time for the IETF to acknowledge the fact that these link types are common and to take a more architectural and wide approach to solving and adapting its protocols to this subnet model.

I'm concerned that we are standardising point solutions without understanding the problem. and I disagree with the chairs consensus call. (not that I necessarily think there are alternatives, but I'd like to see a big warning banner somewhere, not just a "we need to fix a few nits before last calling it).

cheers,
Ole


On Oct 28, 2010, at 23:58 , David Allan I wrote:

> A quick comment on the soapbox statement...
> 
> <soapbox statement>: I'm biased against this subnet model (N:1)... recreating PPP functionality over Ethernet, trying to create user isolation on a shared IPv6 link, which after all IPv6 protocols are not designed for.
> 
> I appreciate the IETF has been kind of blind to this but this kind of asymmetric Ethernet subnet is actually much more prevalent and been around longer than you might think. In Metro Ethernet Forum terms it is known as an E-TREE, support for which is being discussed by the IETF L2VPN WG. IEEE 802.1ad(2005?) documents one possible means of implementing this in the form of Asymmetric VID (which I think is also known as private VLAN) and this has been carried forward into 802.1ah PBB/.1aq SPB. 
> 
> There is also physical media that behaves like this in the form of passive optical networks, which are p2p to the root and broadcast to the leaves. GPON and EPON becoming a very prevalent broadband deployment model.... 
> 
> BBF TR101(2006) is simply one instantiation of a useful construct that has been around for years...and if anything will become much more common over time...
> 
> Cheers
> Dave