[Serusers] How to mangle SDP with result of shellscript

Klaus Darilion klaus.mailinglists at pernau.at
Thu Feb 17 00:09:41 CET 2005


And how does it behave in this situation?


             1.2.3.4
               NAT
              /   \
       10.0.0.2   10.0.0.3
          NAT        NAT
          /            \
      192.168.0.2  192.168.0.2

The clients have the same "public IP address" 1.2.3.4 and the same local 
IP address.

IMO getting back the local IP address. The client peforms NAT traversal 
and you are going to revert the NAT traversal. Even if you can recover 
the IP address from the call-id, how do you find out the local RTP port 
if the client uses stun?

regards,
klaus


Alexander Philipp Lintenhofer wrote:

>> are you sure that it is easy? How do you detect if a UA is NATed twice 
>> (or even more times) or not?
>> Just because two UAs have the same external IP does not mean that they 
>> are in same (local) network.
> 
> 
> I save the original IP-Address in a second location database and compare 
> the network parts of caller and callee in addition to the value of the 
> outbound ip-address. This really does not work if they are natted 3 
> times and the first and third have the same RFC1918 net-class. But i 
> don't mention this case - the probability is really low.
> 
> 
> Nils Ohlmeier schrieb:
> 
>> Hi,
>>
>> On Tuesday 15 February 2005 23:37, Alexander Philipp Lintenhofer wrote:
>>  
>>
>>> Hi there,
>>> I have a question concerning the following situation:
>>> One client makes a call to another client in the same (private) LAN over
>>> a SER located in the public network. It is easy to check if both clients
>>> are behind the same NAT. So one can avoid using the mediaproxy and
>>> therefore the mediastream stays in the local LAN.
>>>   
>>
>>
>> are you sure that it is easy? How do you detect if a UA is NATed twice 
>> (or even more times) or not?
>> Just because two UAs have the same external IP does not mean that they 
>> are in same (local) network.
>>
>> Greetings
>>  Nils
>>
>>  
>>
> 
> _______________________________________________
> Serusers mailing list
> serusers at lists.iptel.org
> http://lists.iptel.org/mailman/listinfo/serusers
> 
> 




More information about the sr-users mailing list