On Fri, 2006-04-14 at 08:13 +0200, Frank Shearar wrote:
> "Paul Kyzivat" <[EMAIL PROTECTED]> wrote:
> > So my conclusion is that if a UA receives a 423 response to a register
> > request, and it chooses to reattempt the register request, it SHOULD use
> > a value consistent with the 423 in the subsequent register request AND
> > in any subsequent refreshes to that registration.
> >
> > Does this seem right?
> 
> Yes, it does. It's probably better for a registration to last slightly
> longer than necessary in the interests of decreasing network traffic and the
> overhead in creating sockets, etc. than for a registrar to have to cope with
> loads of UAs hitting it with REGISTERS that it'll just 423 anyway.
> 
> I don't think it'd be necessary to maintain the registrar's preferred value
> in the UA across reboots, mind you. Or the UA could, from time to time,
> re-register with its (smaller) preferred value, just to see if the registrar
> will now accept the smaller value. (Perhaps the registrar's under much less
> load now, and adjusted its desired minimum value accordingly.)

Both of these seem quite reasonable.

Dale
--- 
interop.pingtel.com -- the public SIP phone interoperability test server

_______________________________________________
Sip-implementors mailing list
[email protected]
https://lists.cs.columbia.edu/cucslists/listinfo/sip-implementors

Reply via email to