That article refers to the previous version of MediaProxy (version 1), which was doing about 200 calls per CPU by relaying packets in the user space.
The current version of MediaProxy (version 2) scales several orders of magnitude higher because the relaying is done directly by the Linux kernel, which can go as fast as your network card can handle. The question how fast is does not really matter nowadays, is too fast for any practical purpose and you can always stack multiple relays to work together for load distribution and resilience. So the assumptions of that article are not valid anymore. Adrian _______________________________________________ Users mailing list [email protected] http://lists.opensips.org/cgi-bin/mailman/listinfo/users
