Bill Fenner wrote:
> On 6/2/06, Joe Touch <[EMAIL PROTECTED]> wrote:
>> Do you recall what was the case that motivates the condition ("IF")
>> regarding how TTLs interact with forwarding on the head and tail of
>> tunnels?
> 
> I was only peripherally involved, but my recollection is that the
> intent was to make a host<->host tunnel look the same as a physical
> point-to-point interface.  The TTL decrement that it refers to is
> simply the TTL  decrement that you would perform during IP forwarding,
> and you're not doing IP forwarding if you originated the packet on the
> tunnel interface.

Agreed - or if you terminate it at the tunnel interface.

> (Talking about this is a little tough without shared terminology for
> some of these concepts.  Steve Deering drew me a picture long, long
> ago which I always wanted to turn into ASCII art for an I-D but never
> did.)
> 
> Given that, the BITW implementation should decrement the TTL if it
> thinks it's a router, or not if it is pretending to be part of the
> host.

Right - I had actually expected that to happen as part of the forwarding
step of the IP payload at the tunnel egress.

Thanks,

Joe

Attachment: signature.asc
Description: OpenPGP digital signature

_______________________________________________
Int-area mailing list
[email protected]
https://www1.ietf.org/mailman/listinfo/int-area

Reply via email to