To:
Robert Elz <kre@munnari.OZ.AU>
cc:
David Terrell <dbt@meat.net>, <ngtrans@sunroof.eng.sun.com>, <namedroppers@ops.ietf.org>, <ipng@sunroof.eng.sun.com>, <dnsop@cafax.se>
From:
Greg Maxwell <gmaxwell@martin.fl.us>
Date:
Thu, 9 Aug 2001 08:03:20 -0400 (EDT)
In-Reply-To:
<944.997327295@brandenburg.cs.mu.OZ.AU>
Sender:
owner-dnsop@cafax.se
Subject:
Re: (ngtrans) Joint DNSEXT & NGTRANS summary
On Thu, 9 Aug 2001, Robert Elz wrote: [snip] > Keith was asking for it to be fixed below the level of the applications, > presumably because updating lots of applications (different kinds, and > just different implementations) is going to be a very long process, > whereas if things could be fixed below that everything would see instant > benefits. That's also why SCTP isn't the immediate answer, as aside > from actually getting that implemented and deployed itself, the applications > would still need to be converted to use it. [snip] Why would this be unacceptable? Having connections survive renumbering is an enhanced feature, and yet another reason applications should move to an improved transport. For the general case of site renumbering, such a change could typically occur in off-hours (where they exist), and go service by service so I don't think it's critical that every application should be able to survive it seemlessly. Mobile IP is a more intresting situation, It would be useful if a mobile host could carry two prefixes, it's tunneled prefix, and a local prefix and could utilize transport layer renumbering support to seemlessly move between networks without the constant overhead of tunneling or dropping connections.