Hi,
Does any vendor support the MPLS for native IPv6 network?
I was working on one testing scenario where I have Cisco 7200 routers and
whole network is running only IPv6 and I wanted to run MPLS on the top of
that but I found MPLS is not supported on IPv6 networks? is that true?
regards
Devang Patel
devang patel wrote:
Hi,
Does any vendor support the MPLS for native IPv6 network?
I think the bigger question is what vendors support native IPv6 networks and at what stage
of maturity.
I was working on one testing scenario where I have Cisco 7200 routers and
whole network is running only IPv6 and I wanted to run MPLS on the top of
that but I found MPLS is not supported on IPv6 networks? is that true?
What sort of environment was this in? A lab environment? Or over a vendors network?
http://www.google.com/search?q=ipv6+mpls&ie=utf-8&oe=utf-8&aq=t&rls=org.mozilla:en-US:official&client=firefox-a might be of interest.
devang patel wrote:
Hi,
Does any vendor support the MPLS for native IPv6 network?
I was working on one testing scenario where I have Cisco 7200 routers and
whole network is running only IPv6 and I wanted to run MPLS on the top of
that but I found MPLS is not supported on IPv6 networks? is that true?
would say it's possible.
However (I'm not an expert at MPLS so forgive me if I'm wrong) isn't MPLS at a lower level and you run IP on top of it? That seems to be what
initial research indicates. See the NANOG talk archives for lots of MPLS presentations.
Hi,
Its lab environment, just created one topology having few routers and
testing it for MPLS VPN for IPv6!!!
the thing is enabling MPLS on interface having IPv6 address will not bring
up LDP neighbor relationship, and due to that it will not assign the MPLS
label to the prefixes!!! so I just want to get information about the same!!!
It was suprising me when I was going to the configuration and
verification!!!
regards
Devang Patel
Unfortunately noone of the major vendors have yet implemented MPLS
control plane via IPv6 transport. From my understanding, the protocol
specs are there, just no implementation.
So for now, you still have to use IPv4 for the MPLS network control
plane and must either forward IPv6 natively dual-stack alongside IPv4,
or transport IPv6 via 6(V)PE.
"no customer demand", as usual. It's just us weirdos trying to do such
things. 
Best regards,
Daniel
Correct, LDP etc. must all be done at IPv4.
That can do the MPLS label bits, and IPv6 can then be carried inside MPLS.
If you use 10.0.0.0/8 for numbering your point to point links, you get 4-8M links, dependant on whether you do /30 or /31.
I don't imagine that represents a problem for many people.
Well, don't know about anybody else, but I've been asking vendors for
LDP6 for a while now, every time we approach them for new projects and
they always look embarrassed when they realise that they
a. don't have it
b. are not involved in the standards process (draft-manral-mpls-ldp-ipv6-02)
So please, if you are spending your hard earned cash, please ask for an
LDP6 implementation, "no demand" should not be the case.
Dave.
Daniel Roesen wrote:
I sent our SE's at C & J yet another reminder about this.
Mark.
<with my vendor hat off...>
If we consider the phases in terms of IPv6 deployment....,
Ph-0: IPv4 only
Ph-1: IPv4/v6 dual stack + v4/v6 coexistence technologies
Ph-2: IPv6 only
I suppose MPLS v6 control plane would become necessary at Ph-2, or a
later stage of Ph-1.
In the meantime, MPLS can be viewed as one of the v4/v6 coexistence
technologies. IPv6 traffic can be carried over v4 signaled MPLS LSP in
various ways (i.e. 6PE/6VPN, static mapping onto TE-LSP, IGP
cutthrough..).
Miya
Hmm, not quite. I'd say:
v4 only
v4/v6 dual stack, with v4 being primary (for network management,
routing, etc)
v4/v6 dual stack, with v6 primary
v6 only
Julien
Hi Julien,
> If we consider the phases in terms of IPv6 deployment....,
>
> Ph-0: IPv4 only
> Ph-1: IPv4/v6 dual stack + v4/v6 coexistence technologies
> Ph-2: IPv6 onlyHmm, not quite. I'd say:
v4 only
v4/v6 dual stack, with v4 being primary (for network
management, routing, etc)
v4/v6 dual stack, with v6 primary
v6 only
"A later phase of Ph-1" in my previous email corresponds to your "v4/v6
dual stack with v6 primary".
And, putting aside whether if it's evil, a certain coexistence
technologies would be needed for the phase transition.
Miya
<snip>
I suppose MPLS v6 control plane would become necessary at Ph-2, or a
later stage of Ph-1.
</snip>
This is a stupidly simple to implement feature, telling us "we're not ready yet"
is not a sensible thing to do, any technology we can have sooner will ease the transition
for many people should they choose to implement it.
I would be interested to know what your views on this are with your vendor hat on
Dave.
I would be interested to know what your views on this are with your
vendor hat on 
Business case conciliates resource contentions...