[afnog] A heads up on a nasty IPv6 bug

Mark Tinka mark.tinka at seacom.mu
Mon Aug 15 09:22:54 UTC 2016



On 15/Aug/16 10:35, Jan Zorz wrote:

> This was the theory, now let's see some practice and real world - what
> can go (and will) go wrong?

Thanks, Jan.

So the main issue is implementation buggy-ness in CPE and client OS
code. Being IPv6, there are many moving parts to make this a quickly
solvable problem in the near future.

Personally, I have no issue hard-coding PD assignments to customers. I
realize IPv4 assignments to customers were random because we were
running out of space in that protocol rather quickly. With IPv6 being as
vast as it is, and considering that customers would generally prefer the
same address space assigned to them for life, it really is a win-win.

I don't run a consumer service network, and the last consumer network I
ran several years ago had not yet started deploying IPv6 to its home
customers, so I can't speak from real experience. However, if I did run
a consumer network, I'd hard-code PD assignments to usernames. It's just
so much easier, and your customers (particularly the tech-savvy ones)
will love you for it! As long as you have a clean mechanism for dealing
with customer churn and space reclamation, you're good.

Mark.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.afnog.org/pipermail/afnog/attachments/20160815/b2b48de1/attachment.html>


More information about the afnog mailing list