Netflix banning HE tunnels

Valdis.Kletnieks at vt.edu Valdis.Kletnieks at vt.edu
Fri Jun 10 08:32:25 UTC 2016


On Fri, 10 Jun 2016 07:19:22 +0100, "tim at pelican.org" said:

> All the business systems that sit around it?  Not so much.  $DAYJOB has
> plenty of code, database structures etc that are built around "an IP address is
> no more than 15 characters long and matches
> '[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}'".  To fix that, you need
> development time - typically expensive analyst time to work out *what* you need
> to change, not just code-grinder make-the-field-bigger time.  IT departments
> seem reluctant to provide that resource, unless you've got people at the very
> top of the business bought into the fact that you *need* to do IPv6.

Some of us had those very same challenges - but managed to deploy IPv6 last
century *anyhow*.  Seriously - none of your code has been untouched for the 16
years so far in this century (if it in fact hasn't, you probably have bigger
problems).  You should have been doing incremental changes for IPv6 as you
revisit code, so you didn't have a big pile of stuff left.

And quite frankly, those of you who dragged their feet have really missed the
boat.  10 years ago, you could have done an incremental deploy of IPv6, knowing
that only people clued enough to manually enable it would be using it.  So
it was safe to put out a "It's sort of there, use at your own risk, let us
know what you manage to break" mode.  SLAAC only because you don't have the
business infrastructure for DHCP6?  Back then it was OK.

That's long gone.  Now every smart phone and laptop - pretty much everything
that isn't a smart TV or a game console is going to try to use it, so you need
to get it 100% right on the first try or your support center is going to catch
fire.



-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 848 bytes
Desc: not available
URL: <http://mailman.nanog.org/pipermail/nanog/attachments/20160610/ddf79b23/attachment.sig>


More information about the NANOG mailing list