On Thu, 21 Jun 2007 11:42:28 -0700
james woodyatt <[EMAIL PROTECTED]> wrote:

> On Jun 21, 2007, at 11:03, Paul Vixie wrote:
> >
> > mark andrews has [observed] that there is no need for the  
> > resolution perimeter of a PTR to differ from the routing perimeter  
> > of the IP address described by that PTR.  yet here we have a large  
> > set of folks who [are] telling us that yes we do need to be able to  
> > resolve a PTR from places where the addresses won't be routable
> > [...]
> > what's wrong with this picture?
> 
> My heretical opinion follows.
> 
<snip>

I agree with James' heretical opinion.

It seems to me that without realising it, we might be discussing
specific instances and coming from different positions on the general
points of 

* the locator and to a slightly lesser extent the identifier properties
of addresses 

* if identifier and locators should come from the same or different
number space, and for different scenarios, how important it is that
they come from the same or different number spaces

* uses of the identifier property only, and the scope of validity for
that identifier **

* uses of the locator property only and the scope of validity for that
locator (e.g. "site", global)

Maybe if we classify the different use scenarios that are being
discussed for ULAs/ULA-Cs using the above points, we'll be in a better
position to evaluate the pro's and con's of ULAs verses ULA-Cs.

Regards,
Mark.

** To explain this better using an example, I think BGPv4 originally
specified that the BGP router ID, carried in the BGP payload, must be
an IPv4 address. I'd be guessing that at the time the key property that
was being sort was BGP router ID uniqueness, and using a global IPv4
address/number value as that ID was a convenient and sure way of
achieving it. I seem to remember reading somewhere that this
requirement has now been relaxed , as it has been recognised that as
long as the BGP router-id is unique within the boundary where
uniqueness is essential, it doesn't have to be a global IPv4 address
(maybe for "pure" IPv6 BGP scenarios, where there wasn't a need to have
the BGP router-id the size of an IPv6 address, but there isn't a global
IPv4 address on the device to assign the router-id value from). It's
certainly still useful to use a global IPv4 address as the BGP
router-id, as if you use the OAM IPv4 address for the device as the BGP
router-id, implying that it is also a locator, it can be very useful for
operational troubleshooting. I think OSPF was less strict on the OSPF
router-id being a global IPv4 address, because OSPF won't ever be
visible globally, however setting the router-id to a global IPv4
address and making that address reachable i.e. an OAM loopback on the
router, is also very common and convenient.

--------------------------------------------------------------------
IETF IPv6 working group mailing list
[email protected]
Administrative Requests: https://www1.ietf.org/mailman/listinfo/ipv6
--------------------------------------------------------------------

Reply via email to