Ah so, this makes sense. I guess you just need to adjust and fine
tune some TCP parameters, take a look at
this,
www.opensips.org/Documentation/Script-CoreParameters-1-8#toc75
[2] Important flags to optimize are,
tcp_connect_timeout
tcp_send_timeout tcp_no_new_conn_flag
tcp_keep
Also as
Yes, we are bridging between TCP and UDP. There are no custom scripts
being used, just the core OpenSIPS scripting, with OpenSIPS acting just
as a proxy. We had another event occur last evening and managed to
capture some interesting information that suggests the problem is
related to blocking
Seems opensips is acting as transport bridge between udp and tcp, so,
do the SIP replies actually arrive at same interface (transport + ip +
port) from where the request was sent, and opensips is listening to that
interface? Can you share SIP trace with us?
Also, are you using any custom appli
Hi all,
We're seeing a possible issue in OpenSIPS related to the timely
processing of replies. For example, OpenSIPS receives a REGISTER and
passes it on to our application server (which is the registrar). The
registrar receives it within the same second and generates the response
(a 401 init