Pekka Savola wrote:
> (Btw, some people didn't know BITW stands for "Bump in the Wire".)
> 
> On Fri, 2 Jun 2006, Joe Touch wrote:
>> FWIW, there are two cases I considered where tunnel decrementing might
>> not occur:
>>
>>     1) BITW
>>         typically this is for IPsec tunnels, which are
>>         spec'd in 4301, but which in spirit ought to follow
>>         2003
>>
>>         they might also be used for range-extenders
> 
> No disagreement about host-to-host tunnels.
> 
> As said, I don't see BITW functionality specified or implied in RFC
> 2003.  But I'd like to know what others think.
> 
> The reason why I think BITW is not important or even relevant in this
> context is that such "BITW-like behaviour" is better achieved by L2
> tunneling, which at the same time can also be agnostic of the various L3
> protocols that might need to be "bumped".

Agreed - I'd like to know what besides 'forwarding' might occur at those
nodes as well.

Come to think of it, L2 bridging is one - where the packet might end up
leaving the device, but the IP tunnel wouldn't decrement the TTL as a
result. That could happen if the tunnel terminates at a VPN router
(e.g., for IPsec transport + IPIP tunnel, as per RFC 3884)

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