[OpenSIPS-Users] Parallel Forking messes up Voicemail two-way audio

[osiris123d]

The call is being set up by the append_branches() and then  
Serialize_branches() with the Q Values being whatever depending on how I  
want the calls to be (parallel or serial). Then in the failure_branch() I  
either call the next_branch() or fail over to voicemail issuing the  
rewritehost() function. One thing I noticed at first is that after the  
serial and parallel calls I get a 482 loop error. So to kind of cheat I  
made my if statement in the failure_branch that catches whether or not it  
fails to VM also catch the 482 message. I am guessing this is not a good  
thing and I need to find out why the 482 loop is occuring.

The other weird thing that I noticed is that if I ngrep on the VM server  
for all traffic on that interface and then test it works!!! How is it that  
just doing an ngrep during the test call would fix the issue??? That is  
what first made me think that a sleep(x) would fix things because I figured  
the ngrep on the VM server was slowing down the VM server.

Anyway I am hoping that the 482 loop issue is the root of my issue. I just  
haven't had time to test more since I am having to study for a CCNP Routing  
test, but I will get back with this issue and send some sip traces unless  
something I said above makes you think what the issue could be.

Thanks

On Nov 16, 2009 8:44am, "Bogdan-Andrei Iancu [via OpenSIPS (Open SIP  
Server)]" <ml-node+4012384-1101982153 at n2.nabble.com> wrote:




> Hi,



> Let me first try to understand the scenario from SIP point of view.



> In the first place, the call is sent to USER and PSTN (parallel). Now,


> the PSTN responds very slow and the USER (callee) decline the call,  
> right ?



> How the VM kicks in? serial forking?



> Maybe a call flow or a trace will help a lot to understand your case.



> Regards,


> Bogdan



> osiris123d wrote:


> > I am wondering if anyone has run into this issue and how it might get  
> fixed.


> >


> > I am testing a Hunt Group call where the user in the location table and  
> a


> > number out on the PSTN both get called at the same time since they both  
> have


> > the same Q value. The parallel forking works just fine but due to the  
> PSTN


> > taking a little longer to respond to the invite the call to the location


> > table user will always cancel before the call to the PSTN number.  
> Because


> > of this I see that the call to the PSTN number is still going on when  
> the


> > Voicemail server picks up. I think because the call to the PSTN user was


> > still in process it messes up the Two-Way audio. You can't hear the  
> audio


> > coming from the Voicemail server. I know for a fact that my mediaproxy


> > functions are set up correctly because on occasion it will work  
> correctly.


> > Any idea how to fix this? I tried the sleep() function in the failure


> > route, but that didn't seem to help.


> >


> >


> --


> Bogdan-Andrei Iancu


> www.voice-system.ro




> _______________________________________________


> Users mailing list


> [hidden email]


> http://lists.opensips.org/cgi-bin/mailman/listinfo/users









> View message @  
> http://n2.nabble.com/Parallel-Forking-messes-up-Voicemail-two-way-audio-tp3925391p4012384.html




> To unsubscribe from Parallel Forking messes up Voicemail two-way audio,  
> click here.








-- 
View this message in context: http://n2.nabble.com/Parallel-Forking-messes-up-Voicemail-two-way-audio-tp3925391p4023033.html
Sent from the OpenSIPS - Users mailing list archive at Nabble.com.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.opensips.org/pipermail/users/attachments/20091117/36969b27/attachment.htm