Summary: | ASTERISK-19825: Bridge stuck for several minutes after hangup - possible hangup control frame skip | ||
Reporter: | Grzegorz Garlewicz (garlew) | Labels: | |
Date Opened: | 2012-05-01 13:28:15 | Date Closed: | 2012-05-15 16:27:47 |
Priority: | Major | Regression? | |
Status: | Closed/Complete | Components: | Core/Channels |
Versions: | 1.8.11.0 | Frequency of Occurrence | Constant |
Related Issues: | |||
Environment: | Attachments: | ||
Description: | In our scenario an outgoing sip call is set up using call file (spool) and after some time bridged to a local extension. At the end the call is effectively sip-local-sip. There is no problem until volume of call grows. Then sometimes after outgoing call hung up the bridge stays not torn down for up to several minutes. I think it might be like one percent calls. I didn't have much time to debug it, what I did so far suggests hangup control frame being queued in ast_queue_frame() but skipped in ast_read(). I will be able to add more debug information next week. | ||
Comments: | By: Matt Jordan (mjordan) 2012-05-04 08:31:41.968-0500 We require a complete debug log to help triage the issue. This document will provide instructions on how to collect debugging logs from an Asterisk machine for the purpose of helping bug marshals troubleshoot an issue: https://wiki.asterisk.org/wiki/display/AST/Collecting+Debug+Information Please also have 'sip set debug on' when generating the log file. By: Grzegorz Garlewicz (garlew) 2012-05-11 11:54:11.985-0500 Ok after some more testing I'm almost convinced that it was fixed in r357761. I was not able to repeat the problem in our lab. We are going to put it in production on monday to confirm it. I did not find this commit because I missed that 1.8.11.0 tag was actually 1.8.11.0-rc1 copy with only a few later commits merged. By: Grzegorz Garlewicz (garlew) 2012-05-15 16:27:13.126-0500 The problem was fixed - bug can be closed now By: Grzegorz Garlewicz (garlew) 2012-05-15 16:27:47.738-0500 Fixed in r357761 |