Summary:ASTERISK-14769: [patch] queue_log is inconsistent for member information - 2 events use member location where the rest use member name
Reporter:warren smith (somedood)Labels:
Date Opened:2009-09-03 16:25:48Date Closed:2011-08-08 15:28:54
Versions:Frequency of
Environment:Attachments:( 0) 20090917__issue15829.diff.txt
( 1) app_queue.c-TRUNK_330273-membername_logging-v9.patch
Description:In 1.6 queue members can now have names, and queue_log records the name (when available, otherwise defaults to location) for member related events except for "ADDMEMBER" and "REMOVEMEMBER".  These events record the agent location instead.

With our queue member setup the location is dynamic depending on where they logged in from, but the name will be static for a member.  With the current queue_log we can't tell when an agent has logged in, but can tell when they've handled calls.  Can ADDMEMBER and REMOVEMEMBER be changed to the same behavior as the other member-related events?


The following queue_log events use membername "RINGNOANSWER", "COMPLETEAGENT", "COMPLETECALLER", "TRANSFER", "CONNECT".

The "ABANDON" and "ENTERQUEUE" events record the string "NONE" for the agent since it doesn't apply.
Comments:By: Leif Madsen (lmadsen) 2009-09-17 14:42:17

This is actually a feature request without a patch provided (which are not handed on the issue tracker).

Additionally, it was pointed out that if this change was made to general Asterisk, we would break existing implementations that require this functionality.

If you need this functionality for your own system, there are many qualified consultants who could implement this on the asterisk-biz mailing list.

By: warren smith (somedood) 2009-09-17 17:28:40

Would it make a difference if I submitted a patch to fix the problem?

We have a workaround setup on our system, but I'm concerned that others that use queue member names without using the agent channel will have the exact same problem.  Can I submit a patch and have the changes included in a changelog/readme or something, and included in the next minor version?

By: Leif Madsen (lmadsen) 2009-09-18 12:08:50

Sure, you can submit it as a patch! :)  I see Tilghman has already set this to Ready for Testing.

Can you round up some testing from the asterisk-users and asterisk-dev mailing list? I'd also like to know if this is a change of functionality that other people require/depend upon.


By: Leif Madsen (lmadsen) 2009-11-04 11:44:52.000-0600

Is the reporter still interested in this functionality? Can you test the patch that was supplied for you?

By: millsu2 (millsu2) 2009-11-10 13:40:31.000-0600

When applying the patch to or hunks 1,3,4 and 11 fail. Hunks 4 and 11 fail on 1.6.2. It patch properly on trunk and I am compiling it to test the functionality, can it be updated for the 1.6.1 or 1.6.2 branch?

By: Leif Madsen (lmadsen) 2009-11-10 15:25:07.000-0600

It was created against revision 219231, which may be a bit old now. The patch would probably need to be updated then for latest 1.6.1.x, but should apply fine against the revision listed above.

By: millsu2 (millsu2) 2009-11-10 18:15:05.000-0600

It works on 229228, except if no agent name is specified then there is a warning:
app_queue.c:5372 ql_exec: QueueLog requires arguments (queuename,uniqueid,membername,event[,additionalinfo])

When no agent name is specified it logs the interface name in place of the agent name in the additionalinfo column of the REMOVEMEMBER event, but that column is blank in the ADDMEMBER event. Also, it seems the agent name should be in the membername column and the interface name should be in the additionalinfo column or not be included at all. None of the other events include that info unless no agent name is specified.

By: Jamuel Starkey (jamuel) 2011-06-24 13:37:04.718-0500

Patch created for asterisk  This patch addresses the original issue reported by providing a queues.conf general configuration option log-member-as-agent. The default value (log-member-as-agent = no) does not alter the existing logging for the ADDMEMBER and REMOVEMEMBER events to maintain existing (buggy) behavior for the sake of backwards compatibility.

By: Jamuel Starkey (jamuel) 2011-06-24 14:57:35.498-0500

Added to review board https://reviewboard.asterisk.org/r/1286/

By: Jamuel Starkey (jamuel) 2011-08-08 13:27:00.364-0500

This is the same patch that has been uploaded to Reviewboard (https://reviewboard.asterisk.org/r/1286).

Please help and test this so it can be merged.

By: Jamuel Starkey (jamuel) 2011-08-08 13:28:47.252-0500

A patch for trunk has been uploaded.  This is the same patch reviewed at https://reviewboard.asterisk.org/r/1286/

Please test so it can be merged.

By: Kinsey Moore (kmoore) 2011-08-08 15:28:54.589-0500

Patch tested and committed.