Bill Manning wrote:
> ... Even if a site uses global scope addresses for its % 
> internal use nodes & applications, a name resolution that 
> includes both % filtered and unfiltered addresses will cause 
> applications that falsely % assume a single address scope to fail. 
> % 
> % Tony
> 
>       This is something else that seems to have crept out of the 
>       woodwork.  Split DNS.  Doe folks really think that this
>       "feature" is going to be required in addition to mandating
>       a functional DNS?
> 
>       If so, what does this say to/about the IAB statement on the
>       requirement for a single DNS context?

As I recall, the IAB statement says there must be a single root for the
'public name space', and I agree. It does not say we must leak
information about filtered addresses outside the scope of the private
network. The requirement to do that is coming from members of the app
community who insist that despite the reality of private use address
space, their right to build 'ignorant apps'R (pass opaque blobs around
without a clue about the validity of the content at the receiver)
supercede the rights of the network manager to get his job done. 

The argument 'passing around blobs doesn't hurt as long as the values
are not ambiguous' ignores the fact that the network manager that put in
the filter may not want information about internal use nodes exposed in
any shape or form, because configuration errors do happen. The DNS (and
for that matter every name resolution mechanism) needs to be aligned
with the reality of the topology that it is handing out locators for.
The DNS as deployed was defined for the unified Internet of 1985, while
the current Internet includes routing filters (addresses with limited
scope). This discrepancy needs to be fixed. The implementation of the
fix may or may not include what we currently see deployed as split-DNS.
Maybe it is time for a complete overhaul (DNSng) to deal with this and
make the system capable of handling the identifier role being discussed
wrt the multihoming problem.

Tony 


--------------------------------------------------------------------
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