IPv6 addressing schemes

Jeroen Massar jeroen at unfix.org
Tue Nov 9 13:53:57 UTC 2004


On Tue, 2004-11-09 at 14:31 +0100, Pierfrancesco Caci wrote:
> 
> 
> Hello,
> do you have suggestions, recommendations, "don't do that"s 
> about how to subnet and assign IPv6 address space to a backbone
> network out of the /32 that we got allocated ?
> Since we are starting from scratch, I'm especially interested in
> hearing about things absolutely to avoid. 
> I was thinking of giving a /48 for each pop (plus a couple of /48 for
> our offices), maybe with some hierarchical structure so that all POPs
> in the same continent are also in the same /something. 

Didn't you have to give any documentation about your network at all to
the RIR from which you received this allocation?

I would expect them to at least as where you have your customers and how
much space you would be giving you. As the design of your network really
completely depends on where you have your network and where you have
your clients and also, don't forget maybe not important for you but it
will be for your followers if you expect your organization to be around
for a while, where they will be in the coming 50 years and longer...

In general you will be giving a /48 per client, if you are not you are
either lying to the RIR's or you are dead sure they only have one link
you could give them a /64. Thus estimate how many clients you will have
at a certain location in the coming X years, times /48, round it up
quite some bit to a nice multiple of 2 so that it nicely fits and tada.

> Pointers to the right words to put in "justfuckinggoogleit" to return
> something useful are welcome too ;-)

"school network administration planning" ?

Greets,
 Jeroen

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 240 bytes
Desc: This is a digitally signed message part
URL: <http://mailman.nanog.org/pipermail/nanog/attachments/20041109/9fad52ce/attachment.sig>


More information about the NANOG mailing list