[Asterisk-Users] IAX2 as an IETF Standard?

[Adam Hart]

Comment below...

Steve wrote:

>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>On Wednesday 24 March 2004 08:45 pm, James H. Thompson wrote:
>  
>
>>No guarantee then when public IPs match that clients are both on same NAT
>>LAN.
>>
>>Client  A 192.168.0.1 ----- NAT Router A --------- NAT Router X with
>>Public IP 123.123.123.123 --- Internet
>>Client  B 192.168.0.1 ----- NAT Router B ---------|
>>    
>>
>
>The thing is that it's all controlled by your gateway configuration. This is 
>where you define where you find what. You must know the IP (or domain name  
>and use DNS) of where the recipient is. If you are calling a local host you 
>must know the IP. If you call an external host you must also must know his 
>internet address. He'd have a redirect in his firewall that would route to 
>his internal machine. You have no need/use of knowing what his internal IP 
>address is. 
>
>I've done all the above in many combinations.
>
>I have one setup on CA and one in FL.
>
>I have had CA call over IP to FL, then fwd the call to a local external land 
>line and call right back in again on another land line. I have called and 
>transferred calls to a local LAN phone as well as over the Internet.
>
>  
>

I can't really follow what you're saying, the above setup is a problem 
with the current IAX. Put simply, when two people are behind the same 
NAT device and the asterisk box is outside this nat, some NAT routers 
can't bridge the calls so the call is forced to continue to route 
through the asterisk box. This is most common cause of compliant of 
latency for the firefly network. Sure SOME routers understand but most 
don't.