On Mon, Jun 9, 2008 at 9:42 AM, Scott Lawrence <[EMAIL PROTECTED]> wrote: > > On Sat, 2008-06-07 at 14:32 -0400, M. Ranganathan wrote: >> > Actually that is not quite correct ( my error for bad explanation ) . >> > Sipxbridge is dependent upon media relay but not vice versa. Does it >> > really matter to you if they run as two independent processes? I dont >> > mind if the configuration has to be read from two different files but >> > when we had this discussion at the start of the project you wanted to >> > combine them into a single file. On the other hand it makes life >> > easier for me if sipxbridge and the media relay run as an independent >> > process. >> >> Sorry... >> >> On the other hand it makes life easier for me and more efficient if >> they do NOT run as independent processes. So long as sipxbridge + >> relay export the right functionality what does it matter? Please feed >> me as many xml files as you wish. > > But isn't it true that the media relay is also used by the proxy for NAT > traversal? >
Correct. > Assume a 2 system HA configuration with no sipXbridge (using either some > other SBC or gateways for PSTN connections), but _with_ the new proxy > NAT traversal support. Is there any reason why each of the 2 proxies > should not have its own co-resident media relay (presumably each media > relay instance would have its own allocated port range at the NAT)? > None. > If you then have such a configuration, and you add a sipXbridge, it > could (with no changes from what we have now) only share the media relay > with one of the two proxies, but both proxies could route calls through > it, correct? > > I would think that the above would argue for separating the media relay > configuration and process management from that of sipXbridge. The 4.0 > sipXsupervisor has a mechanism to configure process dependencies, so it > can take care of not starting a proxy or a bridge that depends upon a > media relay until after it is up if that is needed. I am starting to see some advantages in separating the function but I have one question - how can the single signaling server ( i.e. sipxbridge that talks to ITSP registrars ) know about the various instances of media relay ( i.e. which ones are alive, which ones it can use etc. ) ? I assume there is some configuration parameter that can tell me at least where all the HA instances are running. I can do mutual pinging to figure out liveness after system restart. How can I find this information? There can be only one instance of signaling (that talks to external registrars). So we need a failover mechanism for that. How shall we do this ? Minor: I assume sipXrelay would be a decent name to give such a service. Ranga > > -- > Scott Lawrence tel:+1.781.229.0533;ext=162 or sip:[EMAIL PROTECTED] > sipXecs project coordinator - SIPfoundry http://www.sipfoundry.org/sipXecs > CTO, Voice Solutions - Bluesocket Inc. http://www.bluesocket.com/ > http://www.pingtel.com/ > > -- M. Ranganathan _______________________________________________ sipx-dev mailing list [email protected] List Archive: http://list.sipfoundry.org/archive/sipx-dev Unsubscribe: http://list.sipfoundry.org/mailman/listinfo/sipx-dev
