Hi,
I have the following situation. I have two ser rel_2_0_0 instances one acting as a boundary proxy the other on as internal proxy (registrar, location).
I have the following situation I have a user which has registered two contacs one woth tcp transport the other one with udp.
All messages coming to and outgoing from my network travels through boundary sip proxy.
When internal ser perform lookup_location it get to contacts and the destination set contains two records: Nov 15 11:03:13 sen ser[3673]: route[FORWARD]: destination set: Contact: sip:xlite@192.168.0.164:1221;rinstance=d35f9cda6bd79b0a;transport=TCP, sip:xlite@192.168.0.164:7060
Then internal ser forward message to boundary SIP proxy by changing the destination of the message and with the usage of t_relay(). I see from log that there are two branches involved: Nov 15 11:03:13 sen ser[3673]: branch_route[BRANCH_ROUTE]: <INVITE> ruri:sip:xlite@192.168.0.164:1221;rinstance=d35f9cda6bd79b0a;transport=TCP t:sip:00@voip.rd.touk.pl f:sip:lkl@voip.rd.touk.pl Nov 15 11:03:13 sen ser[3673]: branch_route[BRANCH_ROUTE]: append_hf(P-Warn: forked branch) Nov 15 11:03:13 sen ser[3673]: branch_route[BRANCH_ROUTE]: <INVITE> ruri:sip:xlite@192.168.0.164:7060;rinstance=cfca8523e3747cb0 t:sip:00@voip.rd.touk.pl f:sip:lkl@voip.rd.touk.pl Nov 15 11:03:13 sen ser[3673]: branch_route[BRANCH_ROUTE]: append_hf(P-Warn: forked branch)
both INVITES reaches boudary SIP proxy.
Because first contact was unreachable there was an tcp error: ser[3516]: ERROR: tcp_blocking_connect: poll error: flags 18 Nov 15 11:03:13 sen ser[3516]: ERROR: tcp_blocking_connect: SO_ERROR (111) Connection refused Nov 15 11:03:13 sen ser[3516]: ERROR: tcpconn_connect: tcp_blocking_connect failed Nov 15 11:03:13 sen ser[3516]: ERROR: tcp_send: connect failed Nov 15 11:03:13 sen ser[3516]: msg_send: ERROR: tcp_send failed Nov 15 11:03:13 sen ser[3516]: ERROR: t_send_branch: sending request on branch 0 failed
after which ser invokes sl_reply_error() which results in 477 next hop error reponse message sent to internal ser instance. The problem arises when boudary sip proxy processes second INVITE which correspond to second destination. when boundary SIP proxy receives it doesn't send it to user agent but instead reponses to internal ser with 500 I'm terribly sorry, server error occurred (1/SL) the only thing I see in the boundary SIP proxy log is: ERROR: sl_reply_error used: I'm terribly sorry, server error occurred (1/SL)
Did I miss something here or this is an error? Waiting for Your feedback
Best regards Tomasz