Juha, > -----Original Message----- > From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On > Behalf Of Juha Heinanen > Sent: 11 May 2008 12:57 > To: Christer Holmberg > Cc: [email protected]; Francois Audet; [EMAIL PROTECTED]; > [EMAIL PROTECTED] > Subject: Re: [Sip] draft-ietf-sip-outbound > > Christer Holmberg writes: > > > But, that still doesn't mean that the UA has to register with every > > edge proxy it is aware of. > > i have been fighting this battle in order to make very simple > implementation possible of what i consider a COMMON case. this common > case is the one where a SINGLE HOST has enough processing > power to serve > as a sip proxy/registrar of the user community, such as a good size > enterprise or even a small voip provider. > > the only problem with this single host implementation of > proxy/registrar > is lack of redundancy. this problem could be trivially > solved by adding > a second sip/proxy registrar that could run independently > from the first > if we could assume that each UA would setup flow and register > with BOTH > of these proxy/registrars when configured with an ob proxy set > consisting of two URIs matching these two proxy/registrars. > > so i would like to propose that ob draft mandates UA to set > up flows and > register with AT LEAST TWO ob proxies if ob proxy set is bigger than > one. second paragraph of 4.2.1. Initial Registrations could then go > like this: > > For each outbound proxy URI in the set, the UAC SHOULD send a > REGISTER request using this URI as the default outbound > proxy. If the > set has more than one URI, the UAC MUST send a REGISTER > request using at > least two of them as the default outbound proxy. [JRE] Or alternatively, why not mandate that a UA has to be able to be configured with an outbound proxy set of between 1 and n entries, where n must not be less than two, and that the UA must be able to register with each of the outbound proxies in its configured set? I see no purpose in a UA being able to be configured with 5, say, entries in its set if it is never going to register with more than 2.
John > > this would make the above described trial implementation of > common case > possible and at the same time would support whatever more complicated > scenarios and also the simple scenario where there is only > one ob proxy > in the set. > > -- juha > _______________________________________________ > Sip mailing list https://www.ietf.org/mailman/listinfo/sip > This list is for NEW development of the core SIP Protocol > Use [EMAIL PROTECTED] for questions on current sip > Use [EMAIL PROTECTED] for new developments on the application of sip > _______________________________________________ Sip mailing list https://www.ietf.org/mailman/listinfo/sip This list is for NEW development of the core SIP Protocol Use [EMAIL PROTECTED] for questions on current sip Use [EMAIL PROTECTED] for new developments on the application of sip
