[j-nsp] Juniper and L2TPv3?

[Zvezdelin Vladov]

Dear All,

Thank you for all of your replies,
but what about this?

Searching L2TPv3 at:
http://www.rfc-editor.org/cgi-bin/idsearch.pl

gives this:

draft-ietf-l2tpext-pwe3-ethernet-02 	Transport of Ethernet Frames over
L2TPv3 	 Rahul Aggarwal 	l2tpext 	7-Dec-04

And within the draft:

Rahul Aggarwal
Juniper Networks

Any suggestion?

Zvezdelin Vladov

On Fri, 14 Jan 2005 02:45:08 +0000, Christopher Morrow
<morrowc at ops-netman.net> wrote:
> 
> On Jan 14, 2005, at 2:24 AM, Pedro Roque Marques wrote:
> 
> > Christopher Morrow writes:
> >
> >> Don't have a document reference, but I recall that their answer to
> >> 'we want you to do l2tpv3 because it's standards based' was: "uhm,
> >> use CCC"
> >
> > I would expect the answer to be "GRE".
> >
> 
> Ah-ha, I have forgotten my tunnelings. I do believe you are correct.
> 
> >> to which the reply was: "Thats not standards based" and
> >> they replied: "yea, but it works the same"... it was a circular
> >> arguement with no real answer arrived at :(
> >
> > My understanding is that the design criteria for l2tpv3 was:
> > a) it must not fit M-series max encaps string size.
> > b) it is convinient to vendor X to create a new copy of GRE since a
> > given set of semantics are desirable for an IP tunneling encaps and
> > the existing implementation of GRE on platform Y doesn't have those
> > semantics (desirable: "precomputed encaps" vs undesirable: "double
> > route lookup" on output).
> 
> I believe the reason to want l2tpv3 was related to some state on the
> tunnels, it's been a while, hence my switching gre and CCC :(
> 
> >
> > That is my recollection of the time i was working for vendor X.
> >
> > It is a solution which serves the interests of one particular vendor
> > ramed through the a standards body, by choosing a WG noboby is paying
> > much attention to... after all L2TP is this dial thing.
> >
> > My personal view, only.
> >
> >> If you paid them enough for the devel work and to move it up the
> >> feature tree I'd bet they'd introduce it for you/us/everyone.
> >
> > That is always true... it can be done on the AS PIC (and probably
> > t-series). But not using m-series normal forwarding path (i.e. tunnel
> > Pic).
> >
> > cheers,
> >   Pedro.
> 
>