[Home]

Summary:ASTERISK-22589: shared_lastcall is not reliable with realtime queues
Reporter:hristo (hristo)Labels:
Date Opened:2013-09-26 06:14:14Date Closed:
Priority:MinorRegression?
Status:Open/NewComponents:Applications/app_queue Resources/res_realtime
Versions:1.8.23.0 13.18.4 Frequency of
Occurrence
Related
Issues:
is related toASTERISK-25692 app_queue: shared_lastcall does not work when using realtime queues and Local channels
Environment:Debian Squeeze 64-bitAttachments:( 0) app_queue-debug-shared_lastcall.diff
Description:I have the following setup:
- queues and queue_members are both realtime
- 2 queues: Q1 with wrapuptime=60; Q2 without wrapuptime
- 1 agent, who is in both queues
- shared_lastcall is set to 'yes'

The problem is that shared_lastcall doesn't have any effect unless both realtime queues are already cached by asterisk. As a result the agent sometimes receives calls from Q2 during the 'wrapuptime' period of Q1.

I did find several tickets that describe similar problems, but none of them mentions realtime, that's why I am opening a new one.

I have managed to track the problem by adding a debug message for every time the app_queue iterates over the queues to find the agent and update his last call time. The iteration only seems to be finding *cached* realtime queues, which in turn is causing the problem I have described above.
Comments:By: Rusty Newton (rnewton) 2013-10-04 18:33:02.897-0500

Can you attach your debug patch? That might be useful for whomever works the issue.

By: hristo (hristo) 2013-10-07 04:06:05.268-0500

Uploading the patch as requested. I have marked the diff as "another type of attachment", otherwise Jira won't let me upload it.

It's probably not a real code contribution anyway as it can only possibly help in resolving this issue. However, if this is not acceptable, please delete the attachment.

By: Rusty Newton (rnewton) 2013-10-22 16:34:05.241-0500

That is fine. My intent was that the patch may help others who look into the issue. It is not intended for commit to source, so you marked it correctly. Thanks!