On Fri, Oct 31, 2008 at 11:51 AM, Scott Brim <[EMAIL PROTECTED]> wrote: > On 10/30/08 5:34 PM, William Herrin allegedly wrote: >> If you asked an audience of system administrators whether Vonage and >> Skype would be impacted by fracturing the end-to-end principle with >> carrier grade NAT, I'm not sure how many would correctly identify that >> Skype would be impacted but Vonage would not. I think most would >> incorrectly believe that all VoIP systems would be hurt, not just the >> ones structured in a peer to peer design. > > OK I'm one of them. I thought Skype had STUN-like servers in order to > deal with NATs. CGNs will affect all such servers but it will affect > them equally. No?
Hi Scott, Skype's primary mode of operation tries to fool the NATs on each side into opening the port that the other side will use. Both clients briefly communicate with a server to determine what the external UDP source address and port is likely to be. Then the clients switch communication to each other and hope the ports match. Behind a small NAT with relatively little simultaneous UDP traffic, this works out pretty well. They finally added some servers to relay calls, yes, but they're the last-ditch fallback and I believe they're only available to paying members though I could have that wrong. Having to provide a lot of these relay servers and the bandwidth that goes with them would severely disrupt Skype's business model. With Vonage, as I understand it, the phones talk to the Vonage servers, not directly to the other Vonage phone. Their main problem is keeping the translation alive without wasting too much bandwidth and dealing with the few NATs that don't implement a generic UDP translation due to port filtering or what have you. Unless there are so many Vonage phones behind the NAT that the NAT runs out of sockets to talk to Vonage, they really aren't affected by the size of the network behind the NAT. Regards, Bill Herrin -- William D. Herrin ................ [EMAIL PROTECTED] [EMAIL PROTECTED] 3005 Crane Dr. ...................... Web: <http://bill.herrin.us/> Falls Church, VA 22042-3004 _______________________________________________ rrg mailing list [email protected] https://www.irtf.org/mailman/listinfo/rrg
