Shouldn't *TCP* use an ephemeral source port, and expect replies to be to the source port? Ben Kaufman -----Original Message----- From: sr-users <sr-users-bounces(a)lists.kamailio.org> On Behalf Of Daniel-Constantin Mierla Sent: Tuesday, March 24, 2020 9:01 AM To: Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>rg>; Juha Heinanen <jh(a)tutpro.com> Subject: Re: [SR-Users] about kamailio.cfg NAT test Hello, this is the test to detect devices behind NAT that use STUN, so they discover properly the public IP of the NAT router, but the port allocation is different for STUN and SIP traffic. If you have an asymmetric signalling client, then this test is not useful -- actually nat traversal cannot be done for asymmetric signalling (when client sends from one address (ip+port) and expects traffic back to another address). You haven't included the contact address, but anyhow probably the reply is not expected on port 44717 as it is indicated in Via. The phone expects some natted environment, as it requested "rport" handling (send back the reply to the port from where the request was received). If it wanted asymmetric signalling, then rport is not needed. If it knew it sends and receives to the same port, also rport would not be needed. Maybe this is just baresip approach not to detect local port of the tcp connection, putting a random value in Via and setting rport parameter. You can tune the test for your specific environment/client apps, but for the vast majority of the cases I encountered over the years, this test is the best one to detect clients behind nat that are using stun. Cheers, Daniel On 24.03.20 13:01, Juha Heinanen wrote: -- Daniel-Constantin Mierla -- www.asipto.com www.twitter.com/miconda -- www.linkedin.com/in/miconda _______________________________________________ Kamailio (SER) - Users Mailing List sr-users(a)lists.kamailio.org https://lists.kamailio.org/cgi-bin/mailman/listinfo/sr-users STATEMENT OF CONFIDENTIALITY: The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain confidential or privileged information. If you are not the intended recipient, or the person responsible for delivering the e-mail to the intended recipient, be advised you have received this message in error and that any use, dissemination, forwarding, printing, or copying is strictly prohibited. Please notify AltiGen Communications immediately at either (888)258-4436 or via email to administrator(a)altigen.com, and destroy all copies of this message and any attachments.

I may be misinterpreting this, but RFC3261 section 18.2.2 ( Sending Responses ): - The server transport uses the value of the top Via header field in - order to determine where to send a response. It MUST follow the - following process: - - o If the "sent-protocol" is a reliable transport protocol such as - TCP or SCTP, or TLS over those, the response MUST be sent using - the existing connection to the source of the original request - that created the transaction, if that connection is still open. - This requires the server transport to maintain an association - between server transactions and transport connections. If that - connection is no longer open, the server SHOULD open a - connection to the IP address in the "received" parameter, if - present, using the port in the "sent-by" value, or the default - port for that transport, if no port is specified. If that - connection attempt fails, the server SHOULD use the procedures - in [4] for servers in order to determine the IP address and - port to open the connection and send the response to. The message is constructed by the application, and the TCP source port is determined afterwards, so there's no way that the application can know what that port will be ahead of time. When using TCP, it should be expected that message source port will be different than the port indicated in the Via. With that said, there's nothing wrong here. nat_uac_test(flags) bit value 16 is : * 16 - Test if the source port is different from the port in the "Via" header This tests exactly what it states, so it's not a false positive. It's probably a bad idea to use bit value 16 to check for NAT if using TCP. Ben Kaufman ben.kaufman(a)altigen.com -----Original Message----- From: Daniel-Constantin Mierla <miconda(a)gmail.com> Sent: Tuesday, March 24, 2020 10:47 AM To: Ben Kaufman <ben.kaufman(a)altigen.com>om>; Kamailio (SER) - Users Mailing List <sr-users(a)lists.kamailio.org>rg>; Juha Heinanen <jh(a)tutpro.com> Subject: Re: [SR-Users] about kamailio.cfg NAT test Yes, tcp is typically using ephemeral source port, in the way that is allocated by the tcp stack/kernel, not easy to be specified by the client app. But the client app can get the local (ephemeral) port after connect() using getsockname() and use it to build the SIP message that is going to be sent through the connection. I expect this is the way how (most of the) sip phone apps do it when the port in SIP headers is matching the source port of the connection. Cheers, Daniel On 24.03.20 15:15, Ben Kaufman wrote: -- Daniel-Constantin Mierla -- www.asipto.com www.twitter.com/miconda -- www.linkedin.com/in/miconda STATEMENT OF CONFIDENTIALITY: The information contained in this electronic message and any attachments to this message are intended for the exclusive use of the addressee(s) and may contain confidential or privileged information. If you are not the intended recipient, or the person responsible for delivering the e-mail to the intended recipient, be advised you have received this message in error and that any use, dissemination, forwarding, printing, or copying is strictly prohibited. Please notify AltiGen Communications immediately at either (888)258-4436 or via email to administrator(a)altigen.com, and destroy all copies of this message and any attachments.