[Home]

Summary:ASTERISK-14895: [patch] Dropping frame since I'm still dialing on Zap/... (resp. DAHDi/...) with DIGITAL calls
Reporter:Udo Schacht-Wiegand (udosw)Labels:
Date Opened:2009-09-27 07:53:05Date Closed:2009-12-01 19:00:32.000-0600
Priority:MinorRegression?No
Status:Closed/CompleteComponents:Channels/chan_dahdi
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) 18623-3.call.txt
( 1) debug-1.txt
( 2) transcap_digital_fix.diff.txt
Description:After call is established, there is no audio or rather no data transfer. Full log is filled with
[2009-09-27 14:24:23] DEBUG[18623] chan_dahdi.c: Dropping frame since I'm still dialing on Zap/1-1...
[2009-09-27 14:24:23] DEBUG[18623] chan_dahdi.c: Dropping frame since I'm still dialing on Zap/1-1...



****** ADDITIONAL INFORMATION ******

Looks to me like ASTERISK-14212 (I couldn't reopen this).

Comments:By: Leif Madsen (lmadsen) 2009-09-28 13:42:01

Can you provide some additional information? There is no console output, and no information to go on. You haven't even indicated whether you're using analog or PRI lines, and you haven't indicated which hardware you're using.

By: Udo Schacht-Wiegand (udosw) 2009-09-30 04:45:50

This is about a DIGITAL call from a Bianca Brick VPN ISDN router to another such router. The router is attached to a Xorcom Astribank (BRI) and the call starts there on channel Zap/62. It goes to 747XXXXX (# anonymized). It goes through Asterisk and goes out to the PSTN via a Digium, Inc. Wildcard TE205P.
(See: "Called g1/747XXXXX"). After the call is answered, the "Dropping frame" starts. The data connection is not established, i.e. the ISDN router gets no CONNECT message.

When I try a voice call via the same setup, it works. Voice calls did not work via this setup with 1.4.25 as shown in issue ASTERISK-14212. So I think, ASTERISK-14212 has been fixed for voice calls, but not for data calls.

When I go back to 1.4.23.1, both voice and data calls work via this setup.

By: Leif Madsen (lmadsen) 2009-09-30 10:41:38

Just pinging alecdavis who uploaded a patch for 15205 to make him aware of this issue. Thanks!

By: Alec Davis (alecdavis) 2009-09-30 14:14:48

I might just have access to a couple of BRI isdn routers, and a DIGIUM BRI card.
So could lab it up.
But I'm busy on PRI-64 and on holiday next week, so I can't see me being able to get onto for a while.

re ASTERISK-14212 vinsik never got back to us, as to whether the dialing flag still being set fixed his problem :(

A pri debug of a call setup may help 'pri intense debug span 1' IIRC.

By: Alec Davis (alecdavis) 2009-10-21 14:55:45

udosw: I this still a problem?

By: Udo Schacht-Wiegand (udosw) 2009-10-23 04:37:46

Yes, definitely. I'll try to provide some more debugging asap.

By: Udo Schacht-Wiegand (udosw) 2009-10-23 05:04:33

I uploaded a 'pri intense debupg span 1' of a failing call.

By: Udo Schacht-Wiegand (udosw) 2009-10-27 10:52:08

I found out that the issue is hardware related. We've got it working on a test machine, Intel board, with Asterisk up to 1.4.27-rc1. We have another machine were digital transmission fails from Asterisk 1.4.23.1 upwards. When I go back to 1.4.19.1 on that machine, digital works. All with the same version of Zaptel and Libpri.

We're going to test on different hardware tomorrow. Will let you know asap as we have results.

By: Leif Madsen (lmadsen) 2009-10-27 10:58:51

Thanks!

By: Udo Schacht-Wiegand (udosw) 2009-11-19 06:50:32.000-0600

[Solved]
This was extremely annoying, took me a few weeks to sort this out. In the end: it is NOT hardware related, just a configuration issue:
1) We had a 'txgain = -5.0' on the span involved. Any txgain other than 0.0 stops DIGITAL calls to work.
2) Also any 't' or 'T' needs to be avoided in the Dial command, obviously the channels need to be brigded natively.
Both points do not matter with 1.4.19.1, but with any higher version both things need to be configured properly to get DIGITAL calls through.

Feature request: Ignore these paramteres with digital calls.



By: Leif Madsen (lmadsen) 2009-11-19 10:10:31.000-0600

Could you bring this up on the asterisk-dev mailing list? There isn't much we can do here right now about this, but perhaps this could be considered a regression which would cause the appropriate commands to get added to Asterisk. I'm not sure at this point though.

By: Udo Schacht-Wiegand (udosw) 2009-11-19 10:44:22.000-0600

Ok, will do. I guess you can close this now.

By: Leif Madsen (lmadsen) 2009-11-19 11:17:57.000-0600

Thanks for the feedback! Closing.

By: Michiel van Baak (mvanbaak) 2009-11-23 01:11:26.000-0600

Re-opened as requested by alecdavis on #asterisk-dev IRC channel.

By: Alec Davis (alecdavis) 2009-11-23 01:24:50.000-0600

further testing as requested by jpeeler on IRC on Friday.
Tested with TE110P without echocan module.

It's feasible, that txgain can be other than 0.0, but still digital calls shouldn't be affected.

in chan_dahdi.conf
txgain=-0.1 to txgain=0.1 Unrestricted digital calls setup and connect with data transfer capability.

txgain of +/-0.2 or more, digital calls setup, but no data transfer.

file chan_dahdi.conf:
...
context=incoming
group=0
pridialplan=unknown
prilocaldialplan=unknown
facilityenable=yes
transfer=yes
switchtype=euroisdn
signalling=pri_net
overlapdial=yes
echocancel=yes
echotraining=900
echocancelwhenbridged=no
rxgain=0.0
txgain=0.2
channel => 1-15,17-31
...
<pre>Equipment setup:
 8102        8202       Analog port 2 for each router
 8101        8201       Analog port 1 for each router
 8100        8200       Data number for each router
ROUTER A    ROUTER B
  ||          ||
  BRA         BRA
  ||          ||
[JTEC ISDN SWITCH]     Direct outbound down to E1
       ||    
       E1    
       ||
     TE110P
       G0              Dialling Group
    [ASTERISK]</pre>



By: Alec Davis (alecdavis) 2009-11-23 01:30:58.000-0600

Version numbers: for ~114122
DAHDI Version: SVN-trunk-r7601 Echo Canceller: MG2
libpri version: SVN-branch-1.4-r1331
Asterisk SVN-trunk-r230381M

By: Alec Davis (alecdavis) 2009-11-23 03:01:09.000-0600

uploaded  transcap_digital_fix.diff.txt for asterisk trunk.
sets actual txgain and rxgain to zero for digital calls.

Tested with setup as above in ~114122. Calls setup and transfer data with tested values upto +/- 5.0.

By: Alec Davis (alecdavis) 2009-11-28 00:37:55.000-0600

A question was raised on IRC.

Is it safe to force the tx/rxgain to 0 for any unrestricted digital calls (which can be SPEECH, AUDIO3.1Khz or DATA)?

The concensus was yes, but the best person to answer this would probably be rmudgett.

By: Digium Subversion (svnbot) 2009-12-01 18:50:30.000-0600

Repository: asterisk
Revision: 232090

U   branches/1.4/channels/chan_dahdi.c

------------------------------------------------------------------------
r232090 | jpeeler | 2009-12-01 18:50:30 -0600 (Tue, 01 Dec 2009) | 10 lines

Do not modify the gain settings on data calls.

(The digital flag actually represents a data call.)

(closes issue ASTERISK-14895)
Reported by: udosw
Patches:
     transcap_digital_fix.diff.txt uploaded by alecdavis (license 585)
Tested by: alecdavis

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=232090

By: Digium Subversion (svnbot) 2009-12-01 18:52:51.000-0600

Repository: asterisk
Revision: 232091

_U  trunk/
U   trunk/channels/chan_dahdi.c

------------------------------------------------------------------------
r232091 | jpeeler | 2009-12-01 18:52:51 -0600 (Tue, 01 Dec 2009) | 17 lines

Merged revisions 232090 via svnmerge from
https://origsvn.digium.com/svn/asterisk/branches/1.4

........
 r232090 | jpeeler | 2009-12-01 18:42:58 -0600 (Tue, 01 Dec 2009) | 10 lines
 
 Do not modify the gain settings on data calls.
 
 (The digital flag actually represents a data call.)
 
 (closes issue ASTERISK-14895)
 Reported by: udosw
 Patches:
       transcap_digital_fix.diff.txt uploaded by alecdavis (license 585)
 Tested by: alecdavis
........

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=232091

By: Digium Subversion (svnbot) 2009-12-01 18:57:02.000-0600

Repository: asterisk
Revision: 232092

_U  branches/1.6.0/
U   branches/1.6.0/channels/chan_dahdi.c

------------------------------------------------------------------------
r232092 | jpeeler | 2009-12-01 18:57:01 -0600 (Tue, 01 Dec 2009) | 24 lines

Merged revisions 232091 via svnmerge from
https://origsvn.digium.com/svn/asterisk/trunk

................
 r232091 | jpeeler | 2009-12-01 18:45:18 -0600 (Tue, 01 Dec 2009) | 17 lines
 
 Merged revisions 232090 via svnmerge from
 https://origsvn.digium.com/svn/asterisk/branches/1.4
 
 ........
   r232090 | jpeeler | 2009-12-01 18:42:58 -0600 (Tue, 01 Dec 2009) | 10 lines
   
   Do not modify the gain settings on data calls.
   
   (The digital flag actually represents a data call.)
   
   (closes issue ASTERISK-14895)
   Reported by: udosw
   Patches:
         transcap_digital_fix.diff.txt uploaded by alecdavis (license 585)
   Tested by: alecdavis
 ........
................

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=232092

By: Digium Subversion (svnbot) 2009-12-01 18:59:23.000-0600

Repository: asterisk
Revision: 232093

_U  branches/1.6.1/
U   branches/1.6.1/channels/chan_dahdi.c

------------------------------------------------------------------------
r232093 | jpeeler | 2009-12-01 18:59:23 -0600 (Tue, 01 Dec 2009) | 24 lines

Merged revisions 232091 via svnmerge from
https://origsvn.digium.com/svn/asterisk/trunk

................
 r232091 | jpeeler | 2009-12-01 18:45:18 -0600 (Tue, 01 Dec 2009) | 17 lines
 
 Merged revisions 232090 via svnmerge from
 https://origsvn.digium.com/svn/asterisk/branches/1.4
 
 ........
   r232090 | jpeeler | 2009-12-01 18:42:58 -0600 (Tue, 01 Dec 2009) | 10 lines
   
   Do not modify the gain settings on data calls.
   
   (The digital flag actually represents a data call.)
   
   (closes issue ASTERISK-14895)
   Reported by: udosw
   Patches:
         transcap_digital_fix.diff.txt uploaded by alecdavis (license 585)
   Tested by: alecdavis
 ........
................

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=232093

By: Digium Subversion (svnbot) 2009-12-01 19:00:31.000-0600

Repository: asterisk
Revision: 232094

_U  branches/1.6.2/
U   branches/1.6.2/channels/chan_dahdi.c

------------------------------------------------------------------------
r232094 | jpeeler | 2009-12-01 19:00:31 -0600 (Tue, 01 Dec 2009) | 24 lines

Merged revisions 232091 via svnmerge from
https://origsvn.digium.com/svn/asterisk/trunk

................
 r232091 | jpeeler | 2009-12-01 18:45:18 -0600 (Tue, 01 Dec 2009) | 17 lines
 
 Merged revisions 232090 via svnmerge from
 https://origsvn.digium.com/svn/asterisk/branches/1.4
 
 ........
   r232090 | jpeeler | 2009-12-01 18:42:58 -0600 (Tue, 01 Dec 2009) | 10 lines
   
   Do not modify the gain settings on data calls.
   
   (The digital flag actually represents a data call.)
   
   (closes issue ASTERISK-14895)
   Reported by: udosw
   Patches:
         transcap_digital_fix.diff.txt uploaded by alecdavis (license 585)
   Tested by: alecdavis
 ........
................

------------------------------------------------------------------------

http://svn.digium.com/view/asterisk?view=rev&revision=232094