[Home]

Summary:ASTERISK-00732: T400P not passing more than 3 digits on T1 E&M trunks
Reporter:ulexus (ulexus)Labels:
Date Opened:2004-01-04 01:35:25.000-0600Date Closed:2011-06-07 14:05:27
Priority:MajorRegression?No
Status:Closed/CompleteComponents:Core/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:
Description:Calls through the PRI work fine, but calls destined for either of my two T100Ps result in the T100Ps only receiving three out of the four digits.
This is only present on CVS from the past week or so.
Asterisk v0.50 works fine.

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

I haven't yet done a lot of testing, but watching the console of each side indicates that the T400P thinks it is sending, for instance, "5140", while the T100P thinks it is receiving "514".  
Dialing out to Bell's PRI is fine.
Signalling on both sides of each T1 is em_w.
I haven't got any extensions yet more or less than four, but I will see if I can get that tested.
I also have only updated the Asterisk with the T400P.  The other two sites are still running v.0.50... They are remote and live and not safe/easy to test.
Comments:By: ltropiano (ltropiano) 2004-01-04 13:13:18.000-0600

by any chance is this a SIP Grandstream phone dialing out of the PRI?  See Bug ASTERISK-733.
By: ulexus (ulexus) 2004-01-04 16:33:41.000-0600

No.  All attempts were done from Asterisk-to-Asterisk routing from an external (over PRI) line, from a Digium FXS, from a Digium FXO, or from the DISA application.  They all behave the same.

I'll try a SIP phone when I get into the office on Monday, but I don't think that is the problem.

It seems to be related to the Asterisk-to-Asterisk communication over the T1 with E&M signalling.
By: jrollyson (jrollyson) 2004-01-14 03:45:03.000-0600

Are you using echotraining? If so, you may need to edit the digits.h and increase the DTMF length or disable echotraining.
By: ulexus (ulexus) 2004-01-14 14:57:04.000-0600

This should not be a DTMF problem, as far I can tell.  The first asterisk box received the correct digits from the FXS and says it sends them to the other asterisk box over the T1 card, using E&M signalling.  The second asterisk box only receives three out of four digits.

That said, no, I am not using echotraining.
By: ulexus (ulexus) 2004-01-14 22:31:54.000-0600

From initiating Asterisk machine:
   -- Executing Dial("IAX2[negia@negia]/16384", "Zap/g1/5140") in new stack
Jan 14 23:16:01 DEBUG[262161]: chan_zap.c:1493 zt_call: Dialing '5140'
Jan 14 23:16:01 DEBUG[262161]: chan_zap.c:1546 zt_call: Deferring dialing...
   -- Called g1/5140
Jan 14 23:16:01 DEBUG[262161]: chan_zap.c:3318 __zt_exception: Exception on 34, channel 25
Jan 14 23:16:01 DEBUG[262161]: chan_zap.c:2727 zt_handle_event: Got event Wink/Flash(3) on channel 25 (index 0)
Jan 14 23:16:01 DEBUG[262161]: chan_zap.c:3162 zt_handle_event: Ignoring wink on channel 25
Jan 14 23:16:02 DEBUG[262161]: chan_zap.c:3318 __zt_exception: Exception on 34, channel 25
Jan 14 23:16:02 DEBUG[262161]: chan_zap.c:2727 zt_handle_event: Got event Hook Transition Complete(12) on channel 25 (index 0)
Jan 14 23:16:02 DEBUG[262161]: chan_zap.c:3318 __zt_exception: Exception on 34, channel 25
Jan 14 23:16:02 DEBUG[262161]: chan_zap.c:2727 zt_handle_event: Got event Dial Complete(9) on channel 25 (index 0)
Jan 14 23:16:02 DEBUG[262161]: chan_zap.c:1094 zt_enable_ec: Enabled echo cancellation on channel 25
Jan 14 23:16:02 DEBUG[262161]: chan_zap.c:1113 zt_train_ec: No echo training requested
Jan 14 23:16:03 DEBUG[262161]: chan_zap.c:3318 __zt_exception: Exception on 34, channel 25
Jan 14 23:16:03 DEBUG[262161]: chan_zap.c:2727 zt_handle_event: Got event Dial Complete(9) on channel 25 (index 0)
Jan 14 23:16:03 DEBUG[262161]: chan_zap.c:1078 zt_enable_ec: Echo cancellation already on
Jan 14 23:16:04 DEBUG[147466]: chan_iax2.c:4248 socket_read: Ooh, voice format changed to 2
By: ulexus (ulexus) 2004-01-14 22:32:40.000-0600

From receiving Asterisk machine:
   -- Starting simple switch on 'Zap/1-1'
 == Unknown extension '514' in context 'default' requested
   -- Playing 'ss-noservice'
   -- Hungup 'Zap/1-1'
By: ulexus (ulexus) 2004-01-14 22:34:41.000-0600

The two are connected via T1 with four dedicated voice E&M channels.

Is this something to do with the wink being disregarded... or maybe it is getting a wink too early and cutting off the number transmission?

Any suggestions on further debugging?
By: ulexus (ulexus) 2004-01-14 22:38:46.000-0600

From receiving Asterisk machine: (now with debug on)
-- Starting simple switch on 'Zap/1-1'
DEBUG[213006]: File chan_zap.c, Line 3392 (zt_read): DTMF digit: 5 on Zap/1-1
DEBUG[213006]: File chan_zap.c, Line 3392 (zt_read): DTMF digit: 1 on Zap/1-1
DEBUG[213006]: File chan_zap.c, Line 3392 (zt_read): DTMF digit: 4 on Zap/1-1
DEBUG[213006]: File chan_zap.c, Line 1017 (zt_enable_ec): Enabled echo cancellation on channel 1
 == Unknown extension '514' in context 'default' requested
By: ulexus (ulexus) 2004-01-14 22:39:42.000-0600

Incidentally, the snips above are from the just-released 0.7.1.
By: ulexus (ulexus) 2004-01-14 22:44:57.000-0600

Nope... not wink.  It does the same thing without wink; jsut take out the two wink-related lines.
By: ulexus (ulexus) 2004-01-14 22:56:34.000-0600

Okay.  After playing around a bit:
Three, four, five, whatever number of digits...
It will only successfully communicate N-1 digits.
If I dial "10203040506070809", the other end will receive "1020304050607080".

Also, this is not affecting FXO, FXS, or PRI signalled ports.
only em_w and em of what I have checked so far.
By: ulexus (ulexus) 2004-01-17 13:19:52.000-0600

Further, if I set the two links to use the data channels and IAX2 instead of the four E&M channels, everything works normally.
By: ulexus (ulexus) 2004-01-25 13:41:29.000-0600

Could someone correct the bug summary to be :
"T400P only passing N-1 digits on T1 E&M trunks"

And, uh, change the "Feedback" status.. :)

edited on: 01-25-04 13:26
By: Tilghman Lesher (tilghman) 2004-01-25 15:57:53.000-0600

If you uncomment the line about OLD_DSP_ROUTINES in the central Makefile, make clean install, and restart Asterisk, do the problems persist?
By: ulexus (ulexus) 2004-01-25 22:16:13.000-0600

Yes.  I am using OLD_DSP_ROUTINES.
By: jrollyson (jrollyson) 2004-01-28 00:37:27.000-0600

The new DSP routines were recently fixed in CVS so that they should no longer have fax detection issues - the old DSP routines are known to have issues with E&M, you'll need the new DSP routines enabled if you want E&M.
By: ulexus (ulexus) 2004-01-28 20:14:13.000-0600

Problem is NOT solved.  With latest CVS and without OLD_DSP_ROUTINES, the same problem exists.
By: Brian West (bkw918) 2004-01-28 23:26:39.000-0600

Contact digium about a firmware update for your E400P if applicable.