Hi,

today, i've noticed the following: In the Config of a Border-Proxy (running Kamailio TRUNK), i have the following settings:

listen=1.2.3.4 listen=1234:1234:3::ea

#!define CAPTURE_NODE "sip:[1234:1234:2:3300::3]:5060"

#!ifdef CAPTURE_NODE # Destination, where to send the traffic modparam("siptrace", "duplicate_uri", CAPTURE_NODE) # Trace all traffic modparam("siptrace", "trace_on", 1) modparam("siptrace", "trace_to_database", 0) modparam("siptrace", "trace_flag", FLT_CAPTURE) modparam("siptrace", "hep_mode_on", 1) #!endif

route { [...] sip_capture(); <== This takes 3-4 seconds, if an IPv4 packet is received!!!! [...] }

I do IPv4 to IPv6 on that node, so a message is received on the IPv4 interface and sent using the IPv6 interface to the core-network and the tracing-server. I've noticed, that forwarding the SIP-Message takes awful long time (3-4 seconds), when the tracing-server is IPv6 and the message is actually received on the IPv4-interface. All works fine, if i have an IPv6 message forwarded via IPv6 or if i have an IPv4 message forwarded via IPv4; but not if i mix those two protocols (i haven't tried forwarding IPv6 messages to an IPv4 tracing server).

Before i dig deeper into the code of the siptrace module, does anyone have a clue, why this could be the case????

Thanks in advance, Carsten