Vicidial INCALL and ERI issue
Posted: Tue Oct 09, 2018 8:23 am
Hi All,
Really need some help.
We have 2 vicidial clusters that are driving me nuts.
1 cluster is setup as a manual dial cluster. The other as an Auto dial cluster.
Manual Dial is currently running.
CentOS 7
MariaDB 10.1.20
Vici VERSION: 2.14-683a BUILD: 180809-1547
Asterisk 1.4.39.2-vici
Auto Dialer is running.
CentOS 7
MariaDB 10.2.13
Vici VERSION: 2.14-683a BUILD: 180809-1547
Asterisk 11.25.3-vici
So, different in quite a number of ways.
My problem.
I am seeing that on both these cluster I constantly have 100's of leads on status 'INCALL'. I have been keeping an eye on both of these clusters and I see the dispo 'INCALL' being allocated to agents at about 2 every second... This makes no sense. Not long and basically all the leads are used up and are stuck in either 'INCALL' or eventually 'ERI' when the sessions are closed or agent machines are reboot or whatever else would trigger closing of the sessions.
Now, as mentioned, this is on 2 separate clusters, and this happens on both manual and auto dial(Ratio). I have to constantly refresh the vicidial list statuses, CSLR, call counts and 'QUEUE' just to have calls passing through. I have taken a slow stepped approach at changing one setting at a time, monitoring it, and then changing it back if I don't see any change in behaviour. Have found nothing that improves, changes or impacts this behaviour at all.
The auto dialer has about 50 agents active with 69 trunks, minimum hopper level of 200 and a ratio of 1.1.
The manual dialer has about 6 agent, used by trainees mostly, 69 trunks and further nothing special.
We have replaced a common core switch, have checked multiple agent machines, fly leads, cable runs, wall boxes, patch panels and everything in between. Nothing has solved this.
About a week ago, on a Wednesday morning, the auto dialer managed to dispo 22 000+ leads as INCALL in about an hour (with only 50 agents mind you), and I don't know why. Given that we have a few different clusters, even in other countries, and I have cross referenced the settings etc between the other systems as well and still can't figure out what in heavens name is causing this.
I have read all the pages I can, nothing similar to our setup. I have had the telco check the lines, we have checked the leads, I have changed agent browsers, phones, and who knows what else over the past month, and I am now giving up.
So, is there anyone that can provide some help and advice, or something that can help me get this mess sorted since it is impacting sales etc as well with agents literally running out of leads because everything either end in 'INCALL' or 'ERI' constantly.
Any help, and advice, and checks, anything. Really, just giving in now hahaha.
Looking forward to your input and guidance.
(FYI, we are currently bench testing the latest version for rollout in December.)
Really need some help.
We have 2 vicidial clusters that are driving me nuts.
1 cluster is setup as a manual dial cluster. The other as an Auto dial cluster.
Manual Dial is currently running.
CentOS 7
MariaDB 10.1.20
Vici VERSION: 2.14-683a BUILD: 180809-1547
Asterisk 1.4.39.2-vici
Auto Dialer is running.
CentOS 7
MariaDB 10.2.13
Vici VERSION: 2.14-683a BUILD: 180809-1547
Asterisk 11.25.3-vici
So, different in quite a number of ways.
My problem.
I am seeing that on both these cluster I constantly have 100's of leads on status 'INCALL'. I have been keeping an eye on both of these clusters and I see the dispo 'INCALL' being allocated to agents at about 2 every second... This makes no sense. Not long and basically all the leads are used up and are stuck in either 'INCALL' or eventually 'ERI' when the sessions are closed or agent machines are reboot or whatever else would trigger closing of the sessions.
Now, as mentioned, this is on 2 separate clusters, and this happens on both manual and auto dial(Ratio). I have to constantly refresh the vicidial list statuses, CSLR, call counts and 'QUEUE' just to have calls passing through. I have taken a slow stepped approach at changing one setting at a time, monitoring it, and then changing it back if I don't see any change in behaviour. Have found nothing that improves, changes or impacts this behaviour at all.
The auto dialer has about 50 agents active with 69 trunks, minimum hopper level of 200 and a ratio of 1.1.
The manual dialer has about 6 agent, used by trainees mostly, 69 trunks and further nothing special.
We have replaced a common core switch, have checked multiple agent machines, fly leads, cable runs, wall boxes, patch panels and everything in between. Nothing has solved this.
About a week ago, on a Wednesday morning, the auto dialer managed to dispo 22 000+ leads as INCALL in about an hour (with only 50 agents mind you), and I don't know why. Given that we have a few different clusters, even in other countries, and I have cross referenced the settings etc between the other systems as well and still can't figure out what in heavens name is causing this.
I have read all the pages I can, nothing similar to our setup. I have had the telco check the lines, we have checked the leads, I have changed agent browsers, phones, and who knows what else over the past month, and I am now giving up.
So, is there anyone that can provide some help and advice, or something that can help me get this mess sorted since it is impacting sales etc as well with agents literally running out of leads because everything either end in 'INCALL' or 'ERI' constantly.
Any help, and advice, and checks, anything. Really, just giving in now hahaha.
Looking forward to your input and guidance.
(FYI, we are currently bench testing the latest version for rollout in December.)