Re: [RRG] Are host-stack modifications allowed or disallowed ?

[Bengt Ahlgren <bengta@sics.se>]

Dow Street <dow.street@linquest.com> writes:

> On Mar 6, 2008, at 10:11 AM, Bengt Ahlgren wrote:
>
>> "William Herrin" <bill@herrin.us> writes:
>>
>>> Loose and strict source routing existed in IPv4 for a reason. There
>>> were some security issues, some performance issues and the network
>>> got
>>> wider than they could support but they weren't fundamentally a bad
>>> idea. Indeed, if loose source routing was still valid, our map-encap
>>> schemes might not need the encap part at all.
>>
>> This is IMO an important point.  You can get the same function as a
>> tunnel, but without requiring the other end to decap.  In NIIA we call
>> it routing hint, but it has basically the same function.
>
> You can also avoid certain path MTU issues of tunnels, all while
> making the underlying intentions for the packet more visible /
> explicit.

Yes, but not necessarily, if a router in the middle needs to insert a
new source route.

> I think some folks may equate tunnel endpoints with "a small number
> of specifically configured routers", and source routing with "no
> control because any router could be a source routing way-point".
> This is really a question of the usage model, though, not an inherent
> property of tunneling vs. source routing mechanisms.  I think you
> could use source routing in a controlled way as well, where only a
> small number of specifically configured nodes are configured to act
> as way-points.

Yes, but that is more a configuration issue, not an intrinsic part or
requirement of the mechanism.  And that is exactly why I advocate
source routing before tunneling: the flexibility it gives.

Bengt

--
to unsubscribe send a message to rrg-request@psg.com with the
word 'unsubscribe' in a single line as the message text body.
archive: <http://psg.com/lists/rrg/> & ftp://psg.com/pub/lists/rrg