[asterisk-bugs] [Asterisk 0016789]: [patch] Overlap receiving timeout, plus dialplan latency, causes network to retry SETUP

Asterisk Bug Tracker noreply at bugs.digium.com
Tue Feb 9 17:46:27 CST 2010


A NOTE has been added to this issue. 
====================================================================== 
https://issues.asterisk.org/view.php?id=16789 
====================================================================== 
Reported By:                alecdavis
Assigned To:                
====================================================================== 
Project:                    Asterisk
Issue ID:                   16789
Category:                   Channels/chan_dahdi
Reproducibility:            sometimes
Severity:                   minor
Priority:                   normal
Status:                     ready for testing
Asterisk Version:           SVN 
JIRA:                       SWP-887 
Regression:                 No 
Reviewboard Link:            
SVN Branch (only for SVN checkouts, not tarball releases): 1.6.1 
SVN Revision (number only!): 243988 
Request Review:              
====================================================================== 
Date Submitted:             2010-02-08 19:57 CST
Last Modified:              2010-02-09 17:46 CST
====================================================================== 
Summary:                    [patch] Overlap receiving timeout, plus dialplan
latency, causes network to retry SETUP
Description: 
An overlap received call, should have a PROCEEDING sent when asterisk has
deemed that no further digits are coming, IE when asterisk starts executing
the dialplan.

The issue is, ast_waitfordigit in ss_thread waits 'matchdigittimeout'
which set 3 seconds, then there is dialplan delay, which may take a while
to execute PROGRESS(), PROGRESS() or ANSWER(). In our case, database
lookups.

When the network retries the SETUP, with the same channel and call
reference, libpri deems this as a 'Not a new call' and promptly rejects the
call. Now the call is lost!.

Previously it has required the PROCEEDING() to be the the first line of
the dialplan context.
====================================================================== 

---------------------------------------------------------------------- 
 (0117932) alecdavis (manager) - 2010-02-09 17:46
 https://issues.asterisk.org/view.php?id=16789#c117932 
---------------------------------------------------------------------- 
Below is a capture to our Telco from our Jtec Switch, using overlap
dialling, asterisk isn't involved with this call.

This shows that the TELCO send back a PROCEEDING before the ALERTING, as I
am proposing.<pre>
0395 USER     100210 12:31:43.99 Q.931 SETUP                   C/Ref 0C01 
Org
0397 NETWORK  100210 12:31:44.15 Q.931 SETUP ACKnowledge       C/Ref 0C01
Dest
0399 USER     100210 12:31:45.41 Q.931 INFOrmation             C/Ref 0C01 
Org
0401 USER     100210 12:31:45.91 Q.931 INFOrmation             C/Ref 0C01 
Org
0403 USER     100210 12:31:46.29 Q.931 INFOrmation             C/Ref 0C01 
Org
0405 USER     100210 12:31:46.57 Q.931 INFOrmation             C/Ref 0C01 
Org
0407 USER     100210 12:31:47.06 Q.931 INFOrmation             C/Ref 0C01 
Org
0413 USER     100210 12:31:47.39 Q.931 INFOrmation             C/Ref 0C01 
Org
0418 USER     100210 12:31:47.72 Q.931 INFOrmation             C/Ref 0C01 
Org
0423 USER     100210 12:31:50.96 Q.931 INFOrmation             C/Ref 0C01 
Org
0425 USER     100210 12:31:51.62 Q.931 INFOrmation             C/Ref 0C01 
Org
<b>0429 NETWORK  100210 12:31:52.44 Q.931 CALL PROCeeding         C/Ref
0C01 Dest</b>
0431 NETWORK  100210 12:31:53.38 Q.931 ALERTing                C/Ref 0C01
Dest
0433 USER     100210 12:31:54.86 Q.931 DISConnect              C/Ref 0C01 
Org
0435 NETWORK  100210 12:31:55.03 Q.931 RELease                 C/Ref 0C01
Dest
0437 USER     100210 12:31:55.08 Q.931 RELease COMplete        C/Ref 0C01 
Org</pre> 

Issue History 
Date Modified    Username       Field                    Change               
====================================================================== 
2010-02-09 17:46 alecdavis      Note Added: 0117932                          
======================================================================




More information about the asterisk-bugs mailing list