[jitsi-users] More information on the HOLD problem (includes packet capture)


#1

Yesterday, I posted a problem report concerning an issue I was having
with SIP calls when I placed them on hold. I've still not solved the
problem but I have done a packet capture for the signaling information
for a call where I put someone on hold and the problem manifested. This
capture is all from tcp port 5060.

Can someone have a look and tell me if they see what the problem might
be? Should I capture another port besides 5060?

Thank You,
Anthony

hold_test.pcapng (27.6 KB)

···

--
OpenPGP Key: 4096R/0x028ADF7453B04B15
Other Key Info: http://www.cajuntechie.org/p/my-pgp-key.html
XMPP?Jabber: cajuntech@dukgo.com
VoIP/SIP: 1259010@localphone.com


#2

Yesterday, I posted a problem report concerning an issue I was having
with SIP calls when I placed them on hold. I've still not solved the
problem but I have done a packet capture for the signaling information
for a call where I put someone on hold and the problem manifested. This
capture is all from tcp port 5060.

Can someone have a look and tell me if they see what the problem might
be? Should I capture another port besides 5060?

I don't have time to look at this atm. But: if calls are put on hold, often an intermediate server (e.g. Asterisk) tries to enable music on hold (MoH). This might confuse the clients.

Instead of doing a packet capture, please send Jitsi's logs. These contain samples (1 in AFAIK 5000 packets) of the actual media, which is relevant to debug this. And maybe try a recent nightly. Libjitsi and the SIP stack had some updates that might solve the problem (its unlikely though).

Thank You,
Anthony

Ingo