Call drops after upgrade to 2.8

Dear all,

i noticed that a lot of calls are being dropt, most about 30 seconds in call.
this started to happen when i upgraded from 2.7.x to 2.8

I installed a new machine with elastix (based on freepbx) and i did upgrade that one too to 2.8
and there the same happens.

The stange is that it is not all the calls that are dropping. some calls can stay online for 24+ hours!

see here a log capture of a moment where it went wrong:

I called from 223 to 224:

[Apr 27 16:14:54] VERBOSE[18228] netsock.c: == Using SIP RTP TOS bits 184
[Apr 27 16:14:54] VERBOSE[18228] netsock.c: == Using SIP RTP CoS mark 5
[Apr 27 16:14:54] VERBOSE[18228] app_dial.c: – Called 224
[Apr 27 16:14:54] VERBOSE[18228] app_dial.c: – SIP/224-0000016c is ringing
[Apr 27 16:14:57] VERBOSE[18228] app_dial.c: – SIP/224-0000016c answered SIP/223-0000016b
[Apr 27 16:14:57] NOTICE[18228] rtp.c: Unknown RTP codec 126 received from ‘192.168.1.2’
[Apr 27 16:14:57] NOTICE[18228] rtp.c: Unknown RTP codec 126 received from ‘192.168.1.2’
[Apr 27 16:14:57] NOTICE[18228] rtp.c: Unknown RTP codec 126 received from ‘192.168.1.2’
[Apr 27 16:15:01] VERBOSE[3586] asterisk.c: – Remote UNIX connection
[Apr 27 16:15:01] VERBOSE[18234] asterisk.c: – Remote UNIX connection disconnected
[Apr 27 16:15:08] NOTICE[18228] rtp.c: Unknown RTP codec 126 received from ‘192.168.1.2’
[Apr 27 16:15:18] NOTICE[18228] rtp.c: Unknown RTP codec 126 received from ‘192.168.1.2’
[Apr 27 16:15:28] NOTICE[3634] chan_sip.c: Disconnecting call ‘SIP/223-0000016b’ for lack of RTP activity in 31 seconds
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:1] Macro(“SIP/223-0000016b”, “hangupcall,”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:1] GotoIf(“SIP/223-0000016b”, “1?noautomon”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Goto (macro-hangupcall,s,3)
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:3] NoOp(“SIP/223-0000016b”, “TOUCH_MONITOR_OUTPUT=”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:4] GotoIf(“SIP/223-0000016b”, “1?skiprg”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Goto (macro-hangupcall,s,7)
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:7] GotoIf(“SIP/223-0000016b”, “1?skipblkvm”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Goto (macro-hangupcall,s,10)
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:10] GotoIf(“SIP/223-0000016b”, “1?theend”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Goto (macro-hangupcall,s,12)
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: – Executing [[email protected]:12] Hangup(“SIP/223-0000016b”, “”) in new stack
[Apr 27 16:15:28] VERBOSE[18228] app_macro.c: == Spawn extension (macro-hangupcall, s, 12) exited non-zero on ‘SIP/223-0000016b’ in macro ‘hangupcall’
[Apr 27 16:15:28] VERBOSE[18228] app_macro.c: == Spawn extension (macro-dial-one, s, 37) exited non-zero on ‘SIP/223-0000016b’ in macro ‘dial-one’
[Apr 27 16:15:28] VERBOSE[18228] app_macro.c: == Spawn extension (macro-exten-vm, s, 9) exited non-zero on ‘SIP/223-0000016b’ in macro ‘exten-vm’
[Apr 27 16:15:28] VERBOSE[18228] pbx.c: == Spawn extension (from-internal, 224, 1) exited non-zero on ‘SIP/223-0000016b’
[Apr 27 16:15:29] VERBOSE[18237] manager.c: == Manager ‘admin’ logged on from 127.0.0.1

I hope someone will be able to tell me where to look to find this cause.

Kind regards,

Dear p_lindheimer and RIchard.

Thanks for your input. It pointed me to the correct settings I think.

there is indeed several nat translations in our network setup, and i had in the sip settings NAT on (don’t remember yes or no)

this does not make sense as we have sip connections trough nat and directly.
the failure was not directly only clients on one side of the network so i did not notice that as the problem
but i have changed the setting to route and now the call’s I did test with that always dropped are now open for over 10 minutes.

I will leave those call’s open for the night to see what will happen and monitor the upcoming days closely.

the users of the system told me that they experience first a silence on the line before it drops… does that make sense??

Will keep you updated on the progress.

Kind regards,

Hendrik kroon

PS. I couldn’t find this setting in the older version of freepbx, is this new?

you may also want to check you rtp timer settings in the SIP Settings module.

Hi Tropicalview,

i might be wron here completely but is there a firewall in between? , i looks like the RTP stream closes unexpected.

[Apr 27 16:15:28] NOTICE[3634] chan_sip.c: Disconnecting call 'SIP/223-0000016b' for lack of RTP activity in 31 seconds

i think you need to look with tcpdump/tshark or wireshark to see what happens.
Does it also happens with FPBX 2.7 ? can’t realy imagine that FPBX is the cause.

Hope it helps.

Richard

Calls are dropping every 5 min also I know its RTP but I dont know what settings to change in SIP Settings module.