IPv6 in the enterprise

[Mark Smith]

On Tue, 19 Apr 2011 15:50:19 +0200
"Asbjorn Hojmark - Lists" <lists at hojmark.org> wrote:

> > I don't understand what the benefits are of then using a GUA a
> > default gateway address on the end-nodes is.
> 
> In practice, pinging a link-local address can be difficult for the end-users.
> 
> A Windows user, for example, can't just "ping fe80::1" if he has multiple interfaces (most do). They have to enter something like "ping fe80::1%10" and you won't know in advance which interface number the end-user has to ping. The same goes for ping6 on a Linux machine, where you can't just "ping6 fe80::1" but have to do something like "ping6 -I eth2 fe80::1".
> 
> If you have to guide the end-user through finding the correct link-local interface to ping from, you might as well (and IMO easier) guide them through pinging a GUA, which you know in advance.
> 

Asking them to ping a GUA might be just as or if not harder. Imagine
asking somebody over the phone to type the following in -

ping6 2001:db8:f16d:2366::1


I'm curious in part because I'm pretty sure the IPv6 RFCs specify that
in most cases (BGP being the exception IIRC), next hop addresses are
specified as link-locals. I think you have to be careful if you do
something other than follow the specifications, because while this
other thing might work in the common case that you test, there might be
subtle failure modes that you don't test for because you aren't aware
of them. So in e.g. six months time if you trigger this subtle failure
mode, you'll probably spend a lot of time troubleshooting it because
you won't associate the fault with the decision you made six months ago
to do something different that apparently worked at the time.

> Just my 2 cent.
> 
> -A 
>