Inbound reports not matching

[AlSam]

I have noticed on the inbound/closer reports that the total number of calls which has a 0 length hold time minus the dropped calls with no hold time does not equal the number of calls answered with no hold time. I apologize if this seems confusing. Check out the screenshot for clarification. Notice that under the other groups (5 secs, 10 secs, etc) the numbers match. I looked at the source code for the report page, queried the table (vicidial_closer_log) and found that there are records which have a status of queue. These calls are from the 15th. I think that is whats messing up the numbers. If I update these records with a value of drop in the status fields, the numbers match. Is it safe to go ahead and update all records with queue in the status field as long as they are not from the current day? I apologize if this all seems a bit disorganized as I'm a bit scatter-brained right now. Let me know if there is anything that needs further clarification. Thanks.

[mflorell]

Those stats each measure different things so they will not necessarily add up to the same thing.

[AlSam]

OK. Is there anywhere I can find out with greater detail what those stats measure? Or would you be able to further explain this matter? I find it strange that all the other fields (5 secs, etc) match perfectly every single day. Also, is it normal to have a record that is days old and have a status of queue?

[mflorell]

It would take quite a while to go into detail about what all the possibilities of a call being counted in every one of those fields would be, and I don't have that much free time.

As for QUEUE status, that is usually caused by an agent session going LAGGED and not responding to the call that it is supposed to acknowledge. How many of those do you see?

[AlSam]

In a 15 day period (10/1 to 10/15) there are approx. 300. This represents aprox. 1.7% of all calls (all calls are inbound).

[mflorell]

Are those QUEUE statused calls associated with the same agents or are they distributed across your agents evenly?

[AlSam]

mysql> select user,count(*) from vicidial_closer_log where status in ('QUEUE') and date(call_date)>='2010-10-01' and date(call_date)<='2010-10-19' group by user;
+------+----------+
| user | count(*) |
+------+----------+
| VDCL | 323 |
+------+----------+
1 row in set (0.04 sec)

This is the distribution of calls with queue status.

[mflorell]

This is most likely a network/workstation/load based issue on your system. I checked on two client sites that do 100,000 - 200,000 inbound calls a month and saw only 15 and 8 of these in total for the same time frame.