Agent Dispo status

[ivschakravarthi]

Hi,

Vicdial VERSION: 2.8-403a
BUILD: 130510-1350
asterisk 1.4.21.2
single server
Centos5.5
One telephony card
No cluster


I have upgraded to latest SVN version.
And disabled Agent choosing of disposition after a call because all the agents are staying in disposition mode for longer periods in order to avoid the calls.
Once the call is finished and agent presses hangup button,one more call will be connected after a wrap up time of 1 second.

But in the Real time Report,some times still i can see some agents in DISPO status and staying in DISPO mode for more than 60 seconds.
I have disabled the dispo choose and the agent has to be either in IN CALL or DEAD state but how he is going to DISPO mode again.?
Can you clarify it for me?

Regards,
chak

[williamconley]

go look at the agent screen and look. we don't have any clients who actually use that feature (easier to fire anyone who stays in dispo for more than 15 seconds ... after you fire one loudly the rest have a tendency to NOT violate). trying to modify vicidial to avoid training is like giving your telemarketers crutches and hobbling your manager. you're not doing them any favors.

[ivschakravarthi]

Hi William,

I stay in separate location and my agents stay in some other location so that is not possible to see what on my agent page when issue happens.
Actually what my concern is how the agent is getting showed in DISPO mode in real time report,if dispo page itself is not available in the agent interface.

Regards,
Chak

[williamconley]

Log in as an agent and test it yourself.

Or do what many others do ... get a remote IP camera at the office you are managing.

You could even have a local manager install logmein.com to a computer that has this issue so you can see the screen when it happens. Free software.

[ivschakravarthi]

Hi William,
I found some thing odd in the vicidial_agent_log records table.


mysql> select * from vicidial_agent_log where lead_id='50133'\G
*************************** 1. row ***************************
agent_log_id: 46434
user: 1025
server_ip: 192.168.1.195
event_time: 2013-06-27 12:41:14
lead_id: 50133
campaign_id: SUPER
pause_epoch: 1372329674
pause_sec: 8
wait_epoch: 1372329682
wait_sec: 119
talk_epoch: 1372329719
talk_sec: 0
dispo_epoch: 1372329772
dispo_sec: 53
status: A
user_group: ADMIN
comments: NULL
sub_status: LOGIN
dead_epoch: NULL
dead_sec: 0
processed: N
uniqueid: 1372329718.17410
1 row in set (0.00 sec)

mysql> select * from vicidial_closer_log where lead_id='50133'\G
*************************** 1. row ***************************
closecallid: 50159
lead_id: 50133
list_id: 999
campaign_id: SUPER
call_date: 2013-06-27 12:41:58
start_epoch: 1372329718
end_epoch: 1372329765
length_in_sec: 47
status: A
phone_code: 1
phone_number: xxxxxxxxxxxx
user: 1025
comments: AUTO
processed: N
queue_seconds: 0.00
user_group: ADMIN
xfercallid: 0
term_reason: CALLER
uniqueid: 1372329718.17410
agent_only:
queue_position: 1
1 row in set (0.00 sec)


Observation

In vicidial_closer_log table,the call queue position is 1 and queue seconds is 0 and length of the call is 47 secs which is the talk time of the call.
When coming to vicidial_agent_log table
wait_epoch: 1372329682
wait_sec: 119
talk_epoch: 1372329719
Actual wait value [difference of talk_epoch-wait_epoch i.e., 1372329719-1372329682] is 37 secs wait time but it is updated as 119 seconds and

talk_epoch: 1372329719
talk_sec: 0
dispo_epoch: 1372329772
dispo_sec: 53
Actual talk value is [ difference of dispo_epoch and talk_epoch i.e.,1372329772-1372329719] is 53 seconds is talk value .But the talk value is zero seconds but dispo_sec is updated with 53 seconds.

I have seen around 50 records out of 4000 calls in the above manner.
The record which i shared above is a call at 12:41 PM at which my server don't experience much load also.
What issue is causing the values to swap?


Regards,
chak.

[williamconley]

remind me again how this is related to agent dispo status?

[ivschakravarthi]

I have enabled auto dispo selection,so that agent no need to select a dispo status after the call and he won't be stopped at dispo screen and moves immediately him to the next page.
As i informed in the above discussion,in real time report i can see agent staying in DISPO mode for more than 60 seconds.
You told me to check the agent screen what is there on it when this situation arises.
I checked and what i saw actually is agent is in a CALL but real time report shows it as he was in a DISPO mode.
I refreshed the real time report and clicked on reload now button in order to confirm and it as same and the agent was in call and real time report shows dispo mode.
After that i searched in db with the lead_id reference and i found that swapping of values observation which i have mentioned in the above post.
talk_epoch: 1372329719
talk_sec: 0
dispo_epoch: 1372329772
dispo_sec: 53
Yesterday i was monitoring continuously and around 10 calls came into my notice like this.
Hope you got it.

[williamconley]

Which leads me to a question: How did you install your system? I see you are on CentOS, but not how you installed. It is entirely possible your system is somehow flawed (from the installation).

Additionally: You have upgraded to "the latest" ... when you did this, did you install the sample .conf files during the upgrade?

Have you considered a fresh install to see if this behavior is still present in clean install?