[Home]

Summary:ASTERISK-07371: chan_zap.c is reporting PRI channels in use on ring event, no inbound calls until asterisk restart
Reporter:Paul Tinsley (ptinsley)Labels:
Date Opened:2006-07-20 15:53:41Date Closed:2011-06-07 14:03:06
Priority:MajorRegression?No
Status:Closed/CompleteComponents:Core/General
Versions:Frequency of
Occurrence
Related
Issues:
Environment:Attachments:( 0) console.txt
Description:Within a few hours sometimes a day or so of starting asterisk, inbound ring events on PRI channels are hung up on.  With the following message:
Jul 18 15:45:47 WARNING[24522]: chan_zap.c:8396 pri_dchannel: Ring requested on channel 0/13 already in use on span 1.  Hanging up owner.

I have been on the phone with the Telco monitoring their switch as this happens and they show the channels as idle when the calls are coming in.

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

Platform: CentOS 4.3

A restart of asterisk solves the problem.
related zaptel.conf:
span=1,1,0,esf,b8zs
bchan=13-23
dchan=24

related zapata.conf:
group=1,2
stripmsd=0
callgroup=1
context=inbound
callerid=asreceived
signalling=pri_cpe
pridialplan=unknown
switchtype=national
immediate=no
echocancel=yes
echocancelwhenbridged=yes
priindication=inband
faxdetect=incoming
channel => 13-23
Comments:By: Paul Tinsley (ptinsley) 2006-07-22 15:37:55

Further information, we noticed that the PRI loosing status of channels seemed to have some relation to queue usage.  We changed our queues to use only static members and the problem has yet to surface.  I am not 100% sure it's gone as it has taken a few days for it to crash before.  I will update this next week sometime to confirm or deny this path of a workaround.
By: Serge Vecher (serge-v) 2006-07-24 08:42:09

ptinsley: were you using AgentCallBackLogin at all when this problem occured? If yes, then this is a known issue 6626.
By: Paul Tinsley (ptinsley) 2006-07-24 10:41:21

Yes, I would go ahead and relate this to that one.  It is along the same line, the behavior is not the same but the culprit seems to be the same.  I don't lose the console and other functionality as described in the other, just use of PSTN.
By: Serge Vecher (serge-v) 2006-07-24 10:44:08

was this checked on 1.2.10?
By: Jared Smith (jsmith) 2006-07-27 13:38:19

I'm seeing this same problem 1.2.10, but only using AgentLogin(), not AgentCallbackLogin()... there's definitely something fishy going on here.

I'm running on x86_64, if that makes any difference.
By: Mal Axmo (malaxmo) 2006-07-28 04:04:16

we have the same problem with 1.2.9.1. But we do not use queues!


By: Paul Tinsley (ptinsley) 2006-07-28 09:49:48

I have not tested this on .10 we are standardized on .9 right now, takes awhile to swing to a new release.  I have yet to see the problem come back since I have stopped using agents, all of my call queues now use static members.  So for now that seems to be a valid workaround for my particular scenario.
By: Wayne Hill (meros) 2006-08-04 10:05:24

To chime in with malaxmo we are seeing this occur in 1.2.10 on a system that does not use queues
By: Russell Bryant (russell) 2006-08-05 00:43:43

I have fixed a deadlock in chan_zap in the 1.2 branch that is not yet in any release.  The bug I fixed could have caused a failure such as this.  If you are having this problem, please update to the latest 1.2 branch.

If you have any further issues, then please pursue this issue through Digium technical support (assuming you are using Digium hardware).  The technical support guys will be able to help you narrow down the problem and gather all of the necessary debugging information so that one of the Digium developers can address the issue.

Thanks