Re: Disabling ip4 and IPV6 on F20RC1

[Pavel Simerda <psimerda redhat com>]

----- Original Message -----
> From: "Tore Anderson" <tore fud no>
> To: "Pavel Simerda" <psimerda redhat com>
> Cc: networkmanager-list gnome org
> Sent: Tuesday, December 17, 2013 12:19:44 PM
> Subject: Re: Disabling ip4 and IPV6 on F20RC1
>
> * Pavel Simerda
>
> > 1) First of all it doesn't *specifically* disable kenrel link-local
> > addresses allocation but performs some magic to disable a couple of
> > IPv6 features at once. This wouldn't be a problem in the original
> > poster's case as he wants to disable IPv6 anyway.
>
> As I understand it, link-local addressing is necessary for IPv6 to
> function at all.

Nope. IPv6 can work just as well as IPv4 with manually configured addresses. It's described in the resource I 
linked to. For example OpenVZ (must be used with a patched kernel) does that for its venet virtual networking 
infrastructure. Another example is a ppp links (with normal kernels) don't even distinguish the link-local 
and global addresses. There's actually no reason you would need link-local addresses for static configuration.

> So disabling link-local addressing is analogous to disabling IPv6.

I understand the source of this idea but I believe it is a mistake and source of problems.

> It would be
> like removing the lowest floor of a building and expecting the remainder
> to *not* come crashing down... :-)

The problem with analogies is that they usually don't work as well as pure technical reasoning. Communication 
between two global IPv6 addresses doesn't rely on link-local addresses at all (neighbor solicitations and 
advertisements use the global addresses in that case). Anyone who says something different should back up his 
claim on technical basis.

> > 2) But setting disable_ipv6 doesn't really work as expected. See [1]
> > and especially the note about disable_ipv6 below the table. The truth
> > is that this also wouldn't affect the original poster's use case
> > where the specific interface is (hopefully) expected to be always
> > without IP addresses.
>
> Okay. I'm not sure I've understood what the bug in question is and how
> to tickle it, though.

I think it's described in one of the upstream bug reports but I can't find it right now. The only thing you 
can do if we don't find a description is to try it yourself.

Cheers,

Pavel