[Asterisk-Users] Dynamically setting up/tearing down extensions

[Jeremy McNamara]

DynExtenDB is not even close to being the proper way to achieve dynamic 
extensions.

Karl (klasstek) and myself (mainly Karl) has spent a few clock cycles 
figuring out how to make dynamic extensions happen, but we had no real 
motivation to finish the task.

Find either one of us on IRC or search the mailing list archives.



Jeremy McNamara


Steven J. Sobol wrote:

>Hello, * newbie here,
>
>I'm designing a setup that is to eventually be used in a production 
>virtual PBX/VoIP service.
>
>Customers need to be able to change their setups over the web - I want 
>them to be able to do simple things like setting up call forwarding, as 
>well as more intricate stuff that will require me to re-generate their 
>dialplans. 
>
>Administration of the service is to be web-based.
>
>I'm looking at DynExtenDB (and have played with it). I love that it reads 
>the dialplans out of a MySQL database - that is a critical issue for me. 
>But it has some issues.
>
>I have a test dialplan with one call to Playback() - just plays a wav file 
>then exits. When DynExtenDB() is called, it adds one extra step that calls 
>DynExtenDB_Free()...
>
>--If I let the wav file play to the end, DynExtenDB_Free() is called 
>properly. If I hang up prematurely, it isn't, and it also isn't called if 
>I set the dialplan to dial out (for example, to forward the call to my 
>cell phone).
>
>--If DynExtenDB_Free() *is* called properly, and I then make another call, 
>DynExtenDB() doesn't seem to be called again.
>
>--I'm not sure that setting up a dialplan for extension 'h' is a good 
>idea. What if I call, and then someone else calls and I hang up in the 
>middle of the call? 
>
>I am ready and willing to make changes to the source to DynExtenDB. In 
>fact, I'd like to get it to a point where it could be used in a production 
>environment. But I have a lot of questions before I can do that.
>
>BTW, I have looked in the archives, and it's been suggested that maybe AGI 
>is a better way to handle this sort of thing - but wouldn't the same 
>issues still exist??
>
>Thanks
>   SJS
>
>  
>