Seems to me, that there is a fairly big problem here for which a
solution would be useful, namely, "zeroconf" of small sites (like home
networks). This involves a combination of:
1) It might be useful to generate a ULA automatically, without the
user having to be involved. But, you only want one ULA per site,
which means you need a protocol/practice that handles multiple
routers at the site that think *they* should generate it if no one
else does. And maybe a user wants to disable the use of ULAs. So
you also need a mechanism for saying "don't automatically generate
a ULA on my network"
2) small sites may well have more than one ISP connection, so you have
the problem of automatically (?) identifying the site
boundaries. (Again, presumably, you don't want grandma to be
configuring this stuff, it needs to be automatic.)
3) There may well be more than on link (e.g, ethernet) present, so you
have to subdivide any prefix for the site into individual subnets
and distribute those (sub)prefixes to the appropriate routers. This
is true whether you are using ULAs, global addresses or both. You
also presumably want all the links to be assigned the same subnet
number (for simplicity).
The IETF has never solved the above problem. Yet, if we just leave the
problem to individual vendors, I suspect we will end up with a mess.
Ole Troan <otr...@employees.org> writes:
> the reason for me asking the question was:
> - are these requirements violating any RFC?
no. But this is the sort of thing that I would think is out of scope
for an IETF document to even say, unless it was targeted specifically
at the above problem. So, even though no RFC forbids it, doesn't mean
we recommend it either.
> - as this behaviour is not covered in existing IPv6 RFCs are they
> clear enough for implementors to implement?
Clear, perhaps. But I'm not sure how much it really helps...
> - are these requirements sufficient to solve the whole problem?
Not a chance.
> - do we need any IETF work to cover these 'gaps'?
> - is this a problem which should be solved?
I tend to think that there is work here to be done. And better in the
IETF where a more holistic approach can be taken. But it is a hard
problem and I'm not sure we'd be successful... And we've done zeroconf
work before, without really producing a result.
I do know though that just saying:
> LAN.ADDRESSv6. 3 The device MUST send a Router Solicitation to
> the LAN, to determine if there
> are other routers present.
> MUST
> LAN.ADDRESSv6. 4 If the device determines other routers are
> present in the LAN, and that another
> router is advertising a ULA
> prefix, the device MUST be configurable to
> automatically use this
> information to decide not to advertise its own
> ULA prefix. MUST
> any opinion on these requirements and how they compare with expected
> behavour as specified in RFC4861?
Won't come close to solving your problem and I suspect you will find
that the above doesn't help a whole lot...
Thomas
--------------------------------------------------------------------
IETF IPv6 working group mailing list
ipv6@ietf.org
Administrative Requests: https://www.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------
