[Asterisk-Dev] Uncommon callback

[Paul]

David Woodhouse wrote:

>On Sun, 2005-05-22 at 17:17 +0200, Tamas J wrote:
>  
>
>>What I have to make is:
>>
>>1.the call is routed through PSTN to asterisk1 (#1) which has ISDN PRI
>>interface(s) - leg1
>>2.#1 doesn't pick up the call, neither rejects, it just place into
>>state
>>CALL PROGRESS (?) [maybe gives back alerting tone? probably not]
>>3.#1 notifies that there is a call to number B from number A to system
>>asterisk2 (#2) - through non-voip protocol [xml-rpc, anything else,
>>doesn't matter here]
>>4.#2 dials number B through PSTN - leg2
>>5.#2 dials #1 (on number #1 sent in notification) through SIP - leg3
>>6.#2 interconnects leg2 and leg3 without ISDN CONNECT
>>7.#1 interconnects leg1 and leg2 without ISDN CONNECT
>>8.when B picks up the phone, channels are getting CONNECT message
>>
>>I know that my explanation can be a bit mess, but I wasn't able to
>>write it down better.
>>    
>>
>
>Normally you'd have system #1 forward the call to system #2 itself
>rather than having system #2 call back to #1. It sounds like you're
>trying to work around the fact that system #2 is behind some kind of NAT
>or stupid firewalling. 
>
>If that's the case, you'd do better just to fix that problem at source
>rather than inventing complicated way of working around it. Or if you
>really can't fix it, put up a VPN tunnel between the two machines to
>bypass the firewalling. 
>  
>
I use port forwarding over SSH a lot. It's a lot easier than modifying 
routers and firewalls. There are also things like PPP over SSH. I 
googled up a few helpful pages on that one and am going to try it out 
soon for something I want to do.

To use either of the above for SIP or IAX transport I would do some 
bandwidth testing untunneled and tunneled. I remember seeing some 
comparison charts for vpn methods with bandwidth and latency 
measurements. IIRC - it seemed to indicate some mathods would really 
suck for voip.