On 14.11.2014, at 21.35, Mark Townsley <m...@townsley.net> wrote:
> Our 30 minute time-slot for Routing Protocol Selection discussion turned into
> 80 minutes, and continued in hallways, bars, and an ad-hoc meeting slot the
> following day. This included technical discussion around a partially
> unanticipated requirement for HNCP to support a stub network with a gateway
> that doesn't have sufficient resources to run a routing protocol. A technical
> solution was proposed and discussed ("proposal #1"), using the attached
> slides. "Proposal #2" in the attached slide deck explored one way to support
> "HNCP Fallback" plus a to-be-named "Routing Protocol" at the same time in a
> Homenet. Consensus among participants in the discussion here in Hawaii was
> that Proposal #1 should be adopted. There was strong consensus that we should
> not support Fallback Routing at all in HNCP, hence Proposal #2 was rejected.
At least judging by the slides, #1 is fallback routing with just extra TLV for
RP-running nodes (and apparently the major benefit that it does not call itself
fallback routing). The HNCP-only node has to do fallback-ish routes on anything
it sees on the network anyway (see fallback routing description) or some wild
default route that may or may not work always if HNCP-only node has also other
routes outside the RP ‘cloud’.
> The two currently available routing protocols are IS-IS and Babel but at this
> point there is no consensus on which is preferred, and other protocols may
> emerge during the ongoing discussion. Margaret Wasserman has volunteered to
> work on a draft comparing the two. Factors that might affect the decision
> include (but are not limited to):
>
> - open source implementation maturity, diversity, and portability
I suppose small footprint (both in terms of memory and flash) should be also a
consideration. Given proposal #1’s wild success, everything that does not want
to be strictly leaf/stub has to be able to run this.
Well, looking forward to the minutes.
Cheers,
-Markus
_______________________________________________
homenet mailing list
homenet@ietf.org
https://www.ietf.org/mailman/listinfo/homenet
