Hello Group,
In the ticket 2280 <https://github.com/OpenSIPS/opensips/issues/2280> I was suggested several options how I should handle the issue of NOTIFY not being sent 1. Fix the SIPP scenario - add the local port in the Request-Route header (this is the preferred method) 2. Add a TCP alias for SIPP_PORT -> 5060. This way when the NOTIFY is generated to port 5060, OpenSIPS will find this alias and send it over the same TCP connection 3. Catch the NOTIFY in local_route and reset the tcp_no_new_conn_bflag - this will create a new connection to the UAC_IP:5060 server (note that I haven't tested this, but it should work). Although this might work, in practice, if client is not reachable, the NOTIFY will be lost - it will not be sent over the same connection the SUBSCRIBE created. Regarding option 1 - the actual routing is being done by another Opensips instance, Is there a way to modify the Router header to have the real port? For option 2 - should I do it at the receiving end or at the sender end? Regarding option 3 - how resetting the tcp_no_new_conn_bflag will help? I am not getting even initial NOTIFY which is part of the handle_subscribe function call? Thanks, Vladimir
_______________________________________________ Users mailing list [email protected] http://lists.opensips.org/cgi-bin/mailman/listinfo/users
