Re: [TLS] Proposed text for removing renegotiation

[Martin Thomson <[e-mail]>]

On 28 May 2014 15:25, Geoffrey Keating <[e-mail]> wrote:
> I think this should be handled by TLS, not by having the application
> request rekeying or renegotiation.  If TLS handles it, I don't see why
> there's a need for a special 'renegotiate' or 'change key' message; it
> can quietly change the key when the appropriate limit is hit.

I wasn't proposing that the application be in control of when rekeying
occurs.  (Yes, if an app needs an explicit break in continuity, then
opening a new connection is a perfectly reasonable way to achieve that
goal.)

Yes, you could just roll on through, identify a point where rekeying
is necessary and automatically do it.  That's even more aggressively
spartan than what I've proposed.

A message enables more than just necessary rekeying.  It also allows
for read and write states to be kept in sync.

It also means that you don't have to go to great lengths to contrive a
rekeying scenario in your testing.  I certainly don't want rekeying to
be rare enough that it breaks the first time that it ever actually
happens.  That's a surefire plan for backup generator syndrome.