Hi Hannes, On Mar 28, 2014, at 10:23 PM, Hannes Gredler wrote: > hi stefano, > > On Fri, Mar 28, 2014 at 05:50:16PM +0000, Stefano Previdi (sprevidi) wrote: > | Hi Yakov, > | > | On Mar 28, 2014, at 6:36 PM, Yakov Rekhter wrote: > | > Stefano, > | > > | >> Hi Yakov, > | >> > | >> thanks for your comments. See below. > | > > | > in-line... > | > > | > [clipped...] > | > > | >> The reality of the industry is that we have both dataplanes so > | >> it's obvious that we may want to provide similar functionalities > | >> for both without claiming one technology is the solution for > | >> everything. > | > > | > The reality of the industry is that while we have the MPLS dataplane, > | > we do *not* have the IPv6 data plane that supports IPv6 SR header. > | > | > | well (just to be picky) as a matter of fact we do ;-) > > the real question is what percentage of public Internet routers > > 1) does have MPLS hardware forwarding support > 2) does have IPv6-SR hardware forwarding support.
no, I don't think it's the right question (btw, with such kind of questions e would still have an atm internet). The real question is what is are there operators (service provider or enterprise networks) using ipv6 dataplane and for which an extension to the ipv6 dataplane brings more benefits than constraints. > and > > 3) what is the pricetag for the hardware upgrade. if hardware upgrade is needed. Keep in mind that in many cases it's not required. First because some vendors already support EHs and second because SR does not mandate that all routers are upgrade but rather only the ones that would process the SRH. > furthermore > > 4) what is the cost increase for future hardware to support > a redundant dataplane which does more or less the same thing > than exisiting hardware. cost is an important factor that operators ALREADY takes into account. > | Also, introducing SRH processing in existing v6 networks is not a > | major change (nothing compared to a migration to a different > | dataplane). > > since most v6 hardware implementations i am familiar with, do not have > support for parsing SRH processing, we're probably not familiar with the same implementations... Again, keep in mind that SRH processing is not required in all nodes of the network... think about it. > what you are suggesting here > *is* a migration to a different dataplane. not really. s. > (and i am pretty sure you did not > want to tell the list that introducing support for dataplane #4 on the > internet > is free of charge for all your employers' v6 products). > > /hannes > _______________________________________________ spring mailing list spring@ietf.org https://www.ietf.org/mailman/listinfo/spring
