VoIP Mailing List Archives

[jeff at jeff.net]

Upgrading an ancient customer installation... was running 1.4.23.1 (Trixbox) with Zaptel 1.4.12.9 and a Sangoma A102D, which has been running fine for 5+ years. Customer getting anxious about hardware failure, so we built a new box and installed 1.8.24.0, Dahdi 2.7.0.1, and a new Sangoma A104D. The single active span is an RBS T1 B8ZS/ESF/E&M Wink.

I tried to move one span over one night which was working fine on the old box. Once plugged in there were no alarms, Sangoma wanpipemon utility showed "connected". I tried calling in on a DID number, and in the 'full' log, with debug and verbose set to 100:

[Dec 5 00:51:37] VERBOSE[5283] sig_analog.c: -- Starting simple switch on 'DAHDI/9-1'
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=1.17E+04, Et=1.45E+06, s/n= 0.01
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=2.76E+03, Et=1.10E+06, s/n= 0.00
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=2.06E+04, Et=1.39E+06, s/n= 0.01
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=4.68E+03, Et=1.40E+06, s/n= 0.00
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=0.00E+00, Et=0.00E+00, s/n= -nan
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=0.00E+00, Et=0.00E+00, s/n= -nan
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=0.00E+00, Et=0.00E+00, s/n= -nan
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=0.00E+00, Et=0.00E+00, s/n= -nan
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=0.00E+00, Et=0.00E+00, s/n= -nan
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=1.30E+10, Et=2.11E+12, s/n= 0.01
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: analog_exception 9
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: Exception on 15, channel 9
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: __analog_handle_event 9
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: Got event UNKNOWN/OTHER(131127) on channel 9 (index 0)
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: DTMF Down '7'
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: Begin DTMF digit: 0x37 '7' on DAHDI/9-1
[Dec 5 00:51:38] DEBUG[5283] chan_dahdi.c: Begin DTMF digit: 0x37 '7' on DAHDI/9-1
[Dec 5 00:51:38] DTMF[5283] channel.c: DTMF begin '7' received on DAHDI/9-1
[Dec 5 00:51:38] DTMF[5283] channel.c: DTMF begin ignored '7' on DAHDI/9-1
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=2.02E+10, Et=4.01E+12, s/n= 0.01
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=1.88E+10, Et=3.89E+12, s/n= 0.00
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=4.78E+10, Et=1.17E+12, s/n= 0.04
[Dec 5 00:51:38] DEBUG[5283] dsp.c: tone 1100, Ew=5.10E+03, Et=6.26E+06, s/n= 0.00
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: analog_exception 9
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: Exception on 15, channel 9
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: __analog_handle_event 9
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: Got event UNKNOWN/OTHER(262199) on channel 9 (index 0)
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: Detected digit '7'
[Dec 5 00:51:38] DEBUG[5283] sig_analog.c: End DTMF digit: 0x37 '7' on DAHDI/9-1
[Dec 5 00:51:38] DEBUG[5283] chan_dahdi.c: End DTMF digit: 0x37 '7' on DAHDI/9-1
[Dec 5 00:51:38] DTMF[5283] channel.c: DTMF end '7' received on DAHDI/9-1, duration 0 ms
[Dec 5 00:51:38] DTMF[5283] channel.c: DTMF end accepted without begin '7' on DAHDI/9-1
[Dec 5 00:51:38] DTMF[5283] channel.c: DTMF end passthrough '7' on DAHDI/9-1
[Dec 5 00:51:38] DEBUG[5283] chan_dahdi.c: Enabled echo cancellation on channel 9
[Dec 5 00:51:38] VERBOSE[5283] sig_analog.c: -- Unknown extension '7' in context 'from-pstn' requested
...

At this point I hear 'invalid extension' and get hung up on, but if you grep out all the DTMF events from this call, you get:

root@astsouth:/var/log/asterisk# grep 'DTMF end' /tmp/foo | grep received
[Dec 5 00:51:38] DTMF[5283] channel.c: DTMF end '7' received on DAHDI/9-1, duration 0 ms
[Dec 5 00:51:41] DTMF[5283] channel.c: DTMF end '1' received on DAHDI/9-1, duration 80 ms
[Dec 5 00:51:41] DTMF[5283] channel.c: DTMF end '5' received on DAHDI/9-1, duration 80 ms
[Dec 5 00:51:41] DTMF[5283] channel.c: DTMF end '7' received on DAHDI/9-1, duration 80 ms
[Dec 5 00:51:41] DTMF[5283] channel.c: DTMF end '6' received on DAHDI/9-1, duration 80 ms
[Dec 5 00:51:41] DTMF[5283] channel.c: DTMF end '0' received on DAHDI/9-1, duration 80 ms
[Dec 5 00:51:41] DTMF[5283] channel.c: DTMF end '0' received on DAHDI/9-1, duration 80 ms

And '715-7600' is the DID number I dialed. So why is the new box immediately answering the call before it gets all the ANI digits? I've poured over all Sangoma's docs online for debugging RBS lines - I'm not having any transmission errors, no overruns, and even the audio of 'invalid extension' seems clear before it hangs up on me.

The next maintenance window I swapped out the A104D for a Digium TE122. Exactly the same issue! Some problem with Dahdi then?

Any ideas?

Thanks,

Jeff LaCoursiere
SunFone

[rswagoner at gmail.com]

I have a system with two Sangoma A104D cards running Asterisk 1.8.11-cert10, Dahdi 2.5.0.1, LibPRI 1.4.12, and Wanpipe 3.5.23. The PRI spans are configured with esf,b8zs. Everything has been working great, which is why I haven't updated it further. You might try an older Dahdi version just to see. Although this might be tricky depending on the OS version.

Ryan



On Mon, Dec 9, 2013 at 5:56 PM, Jeff LaCoursiere <jeff@jeff.net (jeff@jeff.net)> wrote:

[rmeyerriecks at digium...]

Looks like DTMF tone generation is too short. That first digit '7'
only lasted 0ms. The rest of the digit tone durations look about right
(80ms). Does this happen on every dial sequence, or just the first?

On Tue, Dec 10, 2013 at 8:35 AM, Ryan Wagoner <rswagoner@gmail.com> wrote: