Andy Pyles wrote:

Hi Jan,

thanks for your answer. however, How does this solution scale?

It scales very well since it was designed from ground-up for performance and robustness. Artifical tests show that on modern hardware it is able to handle some 2,000 simulateneous calls, so that even applying real world discounts you should be able to sustain 1,000 simulateneous sessions without any problems.

I may have a large number of users behind NAT.

Is it possible to have a failover.. i.e. send to to rtpproxy1, if it is down/busy send to rtpproxy2, rtpproxy3, etc? From what I can see in the source code there can only be one rtpproxy at a time. Is this correct? Or am I missing something.

Yes, it is correct, but rtpproxy is quite simple component so that in normal operational conditions there are no or very small probability that it will crash or hang. We have been testing it extensively in production conditions during the last 10 months and never observed any problems due to it. I think that iptel.org also uses it extensively in their service and so far they never reported any stability problems to me.

However, fallover can be relatively easy implemented.

-Maxim

Is anyone deploying a large number of users using rtpproxy? What have you found?

regards, Andy

On Mon, 19 Apr 2004, Jan Janak wrote:

...

Most people are probably using rtpproxy.

Jan.

On 14-04 14:27, Andy Pyles wrote:

...

Hi, I'd like to get some recommendations from other users what the best option for an rtp proxy to use. From what I have seen the options are:

1.) Ser MEDIA proxy

2.) rtpproxy

3.) asterisk ( perhaps with some slight modifications )

What are people using? and any insight would be helpful, as to which model seems to work the best.

regards, Andy

---

Serusers mailing list serusers@lists.iptel.org http://lists.iptel.org/mailman/listinfo/serusers

---

Serusers mailing list serusers@lists.iptel.org http://lists.iptel.org/mailman/listinfo/serusers