Re: delayed multihoming/mobility set-up

[Dave Crocker <dhc@dcrocker.net>]

Folks,

PS>   Is it necessary to ensure connection survivability prior to the connection
PS>   establishment?
PS>   Or, is it necessary to ensure connection survivability in parallel with
PS>   the connection establishment?
PS>   Or even, is it OK to just ensure connection survivability only for
PS>   "long-lived" sessions (e.g., those which have lasted for longer than 5
PS>   minutes), using some definition.

By way of saying the same as the above, but using other words:

The different multiaddressing proposals have very different service
models, in terms of their ability to assure that a connection is
maintained across different locator usage.

If the multiaddressing negotiation precedes connection establishment,
then connection survival is possible for the entire life of the
connection.

If the negotiation is done in parallel with the start of the regular
transport connection, the the early period of the connection will not
survive a change in locator.

If the negotiation is deferred until much later in the life of the
connection, obviously the connection is vulnerable beforehand, but all
of the overhead of the negotiation is avoided for short-lived
connections.

The question, then, is how important people believe this range of
choice and range of service is.


d/
--
 Dave Crocker <dcrocker-at-brandenburg-dot-com>
 Brandenburg InternetWorking <www.brandenburg.com>
 Sunnyvale, CA  USA <tel:+1.408.246.8253>