I need to split split media section and signaling in asterisk code.
Is it possible?
Can you explain your higher level goal as I am having difficulty understanding what you are asking
Your reference to “code” suggests you need the developer mailing list, but the rest of the question doesn’t convince me that that is what you are really after.
The ability and degree to which you can separate media from signalling depends on the device technology being used and which advanced features you require.
As well as things like SIP direct media, there are some facilities to present a different IP address for RTP and SIP, although they both end up on the Asterisk box.
The last two paragraphs make various interpretation of your question that I cannot verify from the question itself.
I designed a hardware for a PBX solution with 2000 analog and sip subscriber.I understand that Asterisk code use high processing power when voice is routed to it. So I need to separate media from Asterisk. In my hardware we use a ARM based processor which support LINUX OS. I decide to run Call control, SIP signaling and some services such as voice mail, web server and databases on this processor. Also I have some DSP processors to handle voice trans-coding and RTP. These DSPs don’t support OS, and I must implement RTP, voice codecs and DTMF detection on them.
Is it possible to use Asterisk in my project?
I need to have all call control PBX features and some Asterisk services on it.
Please guide me.
Thanks a lot.
There is no guide for that sort of environment, you have to determine what capabilities Asterisk has and how to integrate them into your stuff.
You can split up signalling and media by having a SIP proxy in front of your Asterisk machines. Take something like openSIPS and have all your UAs register to that. Now you provision your Asterisk machines as usual, then use some utility like rtpproxy as an offload point on a separate, dedicated host for the actual handling of audio.
To make this work correctly you simply need to configure openSIPS to generate INVITE SDP headers that specify codecs/priorities as well as the new RTP offload host.
In this case Asterisk how find detected DTMF on rtp proxies or Asterisk how can record voice streams for voice mailing service?
Which DTMF method do you use? You can use out-of-band signaling (e.g. SIP INFO) thus completely separating signaling and media streams. Could you provide a high-level design of your setup (including components and traffic flows) to give us understanding about your specific needs?
A new idea come to mind but i don’t know it is possible or not.
Asterisk supports MGCP protocol so it can works with media gateway. I need to have all pbx features, IVR and voice mail services but media must be routed to media gateway. Is it possible?
If i rout media to media gateway, voice mail and IVR how works on asterisk?
chan_mgcp is for having MGCP phones connect to Asterisk… it’s not for having something else handle media for Asterisk.