Re: IPv6 over IPv4 GRE

Christopher,

One of the key diffs is the extra header imposed by GRE between ip header
and encapsulated ip header.

GRE header has 4bytes and support other protocols beyond IP, for example,
IPX and CLNS.

This extra header may cause MTU related issues.

Att.
kratz

On Sat, Aug 28, 2010 at 6:37 PM, Grammer, Christopher <
cgrammer_at_essilorusa.com> wrote:

> All,
>
> I have basic GRE tunnel question.
>
> R3 -- R1 ===GRE=== R2 -- R4
>
>
> Between R3 and R1 is IPv6.
> Between R2 and R4 is IPv6
>
> I am using a manual GRE tunnel between R1 ad R2 which is only configured
> for
> IPv4 1.1.12.0/24
> R1 loopback 1.1.1.1/32
> R2 loopback 1.1.1.2/32
>
> Tunnel12 is IPv6 addressed with with IPv4 source and destination.
>
> RIPng is running between R3 and R1....over the GRE...and between R2 and R4.
> Everything works perfect.
>
> tunnel mode is GRE
>
> And I get debug detail:
>
> *Mar 1 00:24:46.215: IP: s=1.1.1.1 (Tunnel12), d=1.1.1.2
> (FastEthernet0/0),
> len 136, sending, proto=47
>
> Notice protocol 47...which I expect.
> I did not expect the tunnel to work without setting the tunnel mode to
> IPv6IP.
>
> I change the tunnel mode IPv6IP
>
> *Mar 1 00:35:47.387: IP: s=1.1.1.1 (Tunnel12), d=1.1.1.2
> (FastEthernet0/0),
> len 1020, sending, proto=41
> Protocol 41 which I expect.
> Ignore the length I was playing with MTU also.
>
> Tunnel mode IPIP doesnt work.
>
> My question is:
> Is it wrong to use GRE encapsulation in this scenario?
> The only difference I see is that the protocol number is different and it
> doesnt follow an RFC i suppose.
>
> Thanks for your help in advance,
>
> Chris
>
>
> Blogs and organic groups at http://www.ccie.net
>
> _______________________________________________________________________
> Subscription information may be found at:
> http://www.groupstudy.com/list/CCIELab.html
>
>
>
>
>
>
>
>

-- 
"Any fool can know. The point is to understand."
                                                Albert Einstein
Blogs and organic groups at http://www.ccie.net