RE: FW: Next steps discussing Routing Challenges of Semantic Addressing

[Dirk Trossen <[e-mail]>]

Hi Alex,

Please see inline.

Best,

Dirk

-----Original Message-----
From: ipv6 [mailto:[e-mail]] On Behalf Of Alexandre Petrescu
Sent: 06 May 2021 22:22
To: [e-mail]
Subject: Re: FW: Next steps discussing Routing Challenges of Semantic Addressing

[SNIP]

> Or is this so off topic to say ipv6 or 6man lists that it can not be  
> discussed there ?

I think it does make sense to discuss it at IETF.

I doubt though that routing on more bits than a dst address can be qualified as semantics.  Semantics would be if the packets could smell something and routing would happen on that smell.  Or be coloured something, or have a spin, or similar.
[DOT] Semantic is about 'meaning of words'; if you equate 'words' with the (packet) fields taken as input for routing decisions, the meaning of those words/fields matters obviously. Point of the draft is the observation that semantic routing in that sense has been ongoing for a long time; it's a fact of life but worthwhile stating as an observation. Even the routing on IP addresses is semantic routing since the meaning of those addresses matters for many (all?) IP routing solutions. More importantly, the meaning of the IP address field can be changed, particularly in limited domains. Semantic prefixes are one example but also simply using the field to convey an entirely different semantic to route over is possible and has been done. Why is this worth observing now? Maybe it's the ease with which one can 'rummage' around in a packet and use various fields to make routing decisions. If you look at the progression from SDN to P4, for instance, the limitations of SDN to only allow actions on a limited set of fields is all but gone now in P4; the packet is your oyster now, so to speak!


Addressing something with something else than an address also poses a challenge.  On another hand, the semantics of an address could be augmented with port numbers, and then routing could be done based on
address+portnumbers, maybe traffic engineering.

[DOT] This relates to your comment before, i.e., that it makes sense to discuss it at the IETF: I do agree but also agree with Brian that it is eventually the IRTF that may be the target here. The emergence and possibility of semantic routing to be done so easily (see my previous reference to SDN->P4) poses indeed a number of challenges, particularly from an interoperability perspective (which is of particular interest to a standards forum and audience). The insights of the larger (IETF) community are important here as input, while the IRTF may provide the forum to bring together researchers (academic and industrial alike) who develop such semantic routing solutions to exchange views with the wider IETF community on how those solutions all come together and work 'friendly' with each other. This fruitful exchange is something that the IRTF can facilitate. An outcome may be recommendations (to developers of solutions and the IETF) on how semantic routing can/should be done sustainably, i.e., without the challenges that the draft also observes.  

Just some simple thoughts...  I have to read the description first.

Alex

> 
> Thx, Robert
> 
> 
> On Thu, May 6, 2021 at 9:09 PM Manfredi (US), Albert E 
> <[e-mail] <mailto:[e-mail]>>
>  wrote:
> 
> This is something that ought to be of interest to organizations which 
> want to use their own semantics, in the IPv6 prefixes. There was some 
> discussion on 6man some time ago, in which the assumption appeared to 
> be, such a thing is already viable.
> 
> Bert
> 
> -----Original Message----- From: routing-discussion 
> <[e-mail]
> <mailto:[e-mail]>> On Behalf Of Adrian 
> Farrel Sent: Thursday, May 06, 2021 05:49 To: [e-mail] 
> <mailto:[e-mail]>; [e-mail] 
> <mailto:[e-mail]> Subject: [EXTERNAL] Next steps 
> discussing Routing Challenges of Semantic Addressing
> 
> EXT email: be mindful of links/attachments.
> 
> Dear all,
> 
> Back in March we posted an initial draft to open the discussion of 
> research related to routing based on additional meaning assigned to IP 
> addresses, and to routing based on fields other than the destination 
> address field of an IP packet. That draft attracted some  comments, 
> and one thing was clear: we needed to split the draft into  a survey 
> of work that has happened / is happening, and a discussion  of the 
> challenges to routing together with the research question.
> 
> So we have done this now (sorry it took so long) and have posted: 
> https://datatracker.ietf.org/doc/draft-king-irtf-semantic-routing-surv
> ey/
>
>
>
>
>
> 
<https://datatracker.ietf.org/doc/draft-king-irtf-semantic-routing-survey/>
> 
> https://datatracker.ietf.org/doc/draft-king-irtf-challenges-in-routing
> /
>
>
>
>
>
> 
<https://datatracker.ietf.org/doc/draft-king-irtf-challenges-in-routing/>
> 
> 
> These new drafts do not pick up all of the comments received (we 
> focused on the split), but we hope to address the comments in new 
> revisions soon.
> 
> At the same time, we have set up a dedicated mailing list for 
> discussion of "Semantic Addressing Routing and Hardware" (SARAH) at 
> [e-mail] <mailto:[e-mail]>. This is an 
> academic, research-based community where we can discuss all aspects of 
> the drafts, introduce our research to each other, and advertise 
> related conference activity. You can subscribe at 
> https://www.jiscmail.ac.uk/cgi-bin/webadmin?A0=SARAH
> <https://www.jiscmail.ac.uk/cgi-bin/webadmin?A0=SARAH>.
> 
> We hope the new list will help to focus discussion and avoid spamming 
> existing mailing lists.
> 
> Looking forward to hearing your opinions.
> 
> Adrian and Dan
> 
> _______________________________________________ routing-discussion 
> mailing list [e-mail] 
> <mailto:[e-mail]>
> https://www.ietf.org/mailman/listinfo/routing-discussion
> <https://www.ietf.org/mailman/listinfo/routing-discussion>
> 
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list [e-mail] 
> <mailto:[e-mail]> Administrative Requests:
> https://www.ietf.org/mailman/listinfo/ipv6
> <https://www.ietf.org/mailman/listinfo/ipv6>
> --------------------------------------------------------------------
> 
> 
> --------------------------------------------------------------------
> IETF IPv6 working group mailing list [e-mail] Administrative
> Requests: https://www.ietf.org/mailman/listinfo/ipv6
> --------------------------------------------------------------------
> 

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[e-mail]
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------