Hi Keith (and others),

It isn't the purpose of the addressing architecture document to
define the full _use_ of site-local addresses.  The addressing
architecture just defines the prefix that has been assigned for
this purpose...  the full use (along with any restrictions on
that use) will be defined in the scoped addressing architecture
document.

I agree that there is a bit of wording in the addressing architecture
that may need to be changed, if we agree to some sort of restricted
usage model for site-locals.  If necessary, this wording can be
changed in the scoped addressing architecture  an "update" to
the addressing architecture.  Or we can go back and change the
addressing architecture itself once we understand the nature of the
change that we want to make.

At this point, I think that it is most important to figure out
what limitations (if any) the WG can agree (though rough consensus)
to place on the use of site-local addresses.  Then, we can figure
out the exact wording of any required changes to the addressing
architecture.

Margaret


At 08:54 AM 11/14/02, Keith Moore wrote:
>    Site-local addresses are designed to be used for addressing inside of
>    a site which is not permanently connected to the Internet. Using
>    site-local addresses, a subnet ID may be up to 54-bits long, but it
>    is recommended to use at most 16-bit subnet IDs, for convenience of
>    subnet management.

I really don't think this is sufficient.  First because SLs don't work
for a network that has any connection to another network (it doesn't
have to be the public Internet), and second because it says nothing
at all about the problems with SLs.

I suggest

Site-local addresses are designed to be used for addressing inside of
networks which have no connections to external networks.  Other networks
need to provide global addresses to their nodes.  Use of site-local
addresses on networks with external connectivity (especially in the
absence of global addresses) is known to cause problems for some
applications, and such use is currently not recommended.  This recommendation
may be changed once the implications of application use of site-locals
are better understood.  When using site-local addresses, a subnet ID may
be up to 54-bits long, but it is recommended to use at most 16-bit subnet
IDs, for convenience of subnet management.

--------------------------------------------------------------------
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]
--------------------------------------------------------------------
--------------------------------------------------------------------
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