Forget the previous error, it is not related to the issue.
The issue is that when OpenSIPS sends the BYE to the SBC, it opens a new TCP connection to the SBC but uses a randomly chosen source port, therefore we don't see the SIP traffic using ngrep since we are filtering using port. Is there a way to force OpenSIPS to always use port 5060 as source port when opening TCP connections? From: Mickael Marrache [mailto:[email protected]] Sent: Wednesday, May 06, 2015 10:15 AM To: [email protected] Subject: TCP connection closed after the call is established Hi, We systematically get the following error with a particular SBC when a BYE is received: Apr 30 21:41:47 [11535]: ERROR:core:tcp_read: error reading: Connection reset by peer Apr 30 21:41:47 [11535]: ERROR:core:tcp_read_req: failed to read The components present in the call flow are: UAC <-> SBC <-> OpenSIPS <-> UAS The call works fine until a BYE is sent by the UAS to OpenSIPS which doesn't forward the BYE to the SBC. I know the TCP connection between the SBC and OpenSIPS is closed after the call is established, and before the BYE is received. However, I expect OpenSIPS to open a TCP connection with the SBC so that the BYE can be forwarded. Any idea? Thanks, Mickael
_______________________________________________ Users mailing list [email protected] http://lists.opensips.org/cgi-bin/mailman/listinfo/users
