This seems like a good start.

I think it would help to provide some advice to applications regarding
site-locals (and for that matter link-locals).  e.g.

- don't use site-locals in referrals unless you have no global addresses.
- use global addresses in preference to site-locals when opening new 
  connections.

I think there will be enough desire to use site-locals at least for
limited-purposes in networks that also have stable globals, and that
it will be useful to provide some advice on how to constrain such use.

I think there is a need for sites to acquire globals even if they do
not connect to the public Internet.  I don't think coordinating use
of site-locals will scale.

I think it needs to be explained that site-locals don't really provide 
a security benefit.

What did I miss?

> So, lets summarise again:
> 
> - In isolated and unconnected networks, site locals work fine.
> 
> - If a stable global prefix is available, we strongly recommend using that
> and not using site locals.
> 
> - Site locals and global addresses can exist in parallel on the same
> network, but this is likely to cause address selection problems for
> applications.
> 
> - A site is demarcated by the presence of routers that will not forward
> packets using site-local addresses (in source or destination).  Routers that
> are not part of a site should not forward such packets.
> 
> - If multiple networks are to be connected using site local addresses, they
> must be configured to be a single site.
> 
> - Site local addresses are explicitly invalid outside their site.
> Connectivity external to a site MUST NOT be done with site local addresses.
> Sites MUST NOT route packets with site local addresses outside the site.
> 
> - Proxying or otherwise attempting to connect site local addresses to the
> global internet (eg NAT) is explicitly discouraged.
> 
> Note: It might be possible to have site local traffic sent over a tunnel (eg
> a VPN).  In such a situation, the VPN should be treated virtually as part of
> the site.  Site local addresses must not leak outside the tunnel.  Though
> there are probably better ways to do this than using site locals.
> 
> 
> Does this seem a fair summary of principles thus far?
--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to