Re: [DNA] DNAO format revision?

[Greg Daley <greg.daley@eng.monash.edu.au>]

Hi Erik,

Erik Nordmark wrote:
> Greg Daley wrote:
> 
>> Please respond indicating if you like any of the
>> existing proposals (Current, Variant 1 or Variant 2),
>> or prefer something different.
> 
> 
> I think either one is an improvement over the current one in the draft. 
> Whether we want the smaller #2 option (with a bit more complexity) 
> depends on how important we think it is to minimize the size of the DNAO.
> 
> The DNAO is only sent by a routers when not all routers on the link are 
> configured to advertise the same set of prefixes. However, RFC 2461 does 
> allow that the routers advertise different sets of prefixes. I'd suspect 
> that in all well-run IPv6 networks the routers advertise the same set of 
> prefixes (so that they all can be backups for each other). So optimizing 
> the size of the DNAO might not be terribly important.
> But I don't have a problem with format #2 - just questioning the 
> necessity of minimizing something which might not be used much in the 
> real world (but is mostly there to deal with the - perhaps unnecessary - 
> generality in RFC 2461).

I understand that this is not the normal case for IPv6 networks.  My
testbed is currently set up to handle this case though, as one of
the tricky cases in DNA.

I'm not sure if it's terribly important to optimize, other than seeing
a few zeros in the Variant 1 which I implemented before trying Variant
2 (since #1 was slightly easier).

Both work though, and I'm happy to go with either.

What's probably most important is that it works with mismatched
(even if poorly configured???) prefixes, rather than worrying about
the size.

If we want to follow the 'keep it simple' principle, then we could
probably choose Variant 1, but keep track of the alternatives in the
issues list.

Does this sound like a good idea?

Greg