Looking at your production traces I see that sipx is sending multiple request to your Cisco endpoint.
For example there are 2 invites sent in a row. Then later on the ACK is sent twice. Looks like a good old fashion network issue with packets not arriving in a timely fashion. Your "test" traces do not show the duplicate packets. Sipx typically does this when he doesn't receive a respond to the invite in a timely fashion. A wireshark capture at both ends is the best way to see if its a latency issue as you can compare the times of when sipx sent the packets and how long it took the cisco to receive or how long it took it to respond. -M >>> On 6/15/2011 at 09:24 AM, in message <0a1fdb5daa23564f8758ba05d26dcd74fc6...@exchange.telecats.nl>, "Henry Dogger" <h.dog...@telecats.nl> wrote: Hi all, I noticed some lag on our production system, when picking up the phone. When I answer the phone it takes a second before the other side can hear me… I tried to reproduce on our test environment but there was only a slight lag (0.2s) from the phone. I made traces on both systems, and I noticed a difference… On my production system the ACK after the OK is held by the proxy for half a second, apparently the phone does not get this, and the proxy sends another one half a second behind the first attempt. I do not see this behavior on our test system. I attached the traces, does anyone know what could cause this? Using sipXecs-4.2.1 both systems are setup identically. Kind regards, Henry Dogger Telecats BV
_______________________________________________ sipx-users mailing list sipx-users@list.sipfoundry.org List Archive: http://list.sipfoundry.org/archive/sipx-users/
