Skill based Queue assigns the call to wrong agents

[Ronin]

Hello,

3cx version: 16.0.507
Server OS: Debian 9 locally hosted
IP phone: Yealink t42s 66.84.0.95
Provisioning Method: Local
Patton Gateway SN4960/4E60V
Firewall checker passed: Yes

We are experiencing problems in our queues.
We use the "skill based routing Ring all" and some times only the agents in the 3rd or 4th skill level get the calls or just one person from the 3rd skill level.
The agents in the 1st and 2nd lvl skill are available and logged in when this happens.
Is this a known issue?

 

[eddv123]

Firewall checker passed: Yes

Although this sounds to be an issue with call distribution, in this event it does not actually matter if the firewall checker passes since your inbound routes are coming in via the ISDN gateway not SIP trunks.

Even if this were a remote gateway to hosted 3CX you should be using a VPN tunnel and thus be LAN to LAN anyway.

In regards to the issue itself I would look at the binary log viewer in VERBOSE mode for a better understanding of what is going on with the call flow: https://www.3cx.com/docs/3cx-log-viewer/

 

[Ronin]

Hello,

I have checked the 3cx logs (medium logging) and the call is going to the 3rd and 4th skill levels simultaneously.
Is there a setting in 3cx that polls two skill groups?

 

[Ronin]

We have 25 agents and 5 queues (skill based Ring all).
Every agent is in every queue in different skill levels.
Could this be creating a problem in polling the agents?
While agents in skill 1 are available and logged in the queue, the call goes to the 3rd and 4th skill level.
This does not happen all the time.
In the binary log viewer I see the wrong agents being polled.
Most of the agents use the 3cx webclient.

 

[JohnS_3CX]

Every agent is in every queue in different skill levels.
While agents in skill 1 are available and logged in the queue, the call goes to the 3rd and 4th skill level.

Since they belong to multiple skill based queues, is there a chance they were being polled by another queue at the time? If the system thinks they are not available or wrapping up from a previous call, they may be skipped. Wrap up time perhaps?

 

[Ronin]

Hello,

They were not being polled by another queue at the time.
Also wrap up time is set to 2 sec and the agents didn't have any calls earlier.
Each time we have this error I get this information in the 3cx log viewer under verbose mode:

18-05-2020 11:23:43:343 | 9 | L:36183.3[Extn:205]: Terminating targets, reason: SIP ;cause=200 ;text="User hangup"

This error comes after 3cx tries to poll the correct agents. ( Agents do not hang up the call)
Then after this error, the next skill level agents are being polled.
The agents are available and logged in the queues but sometimes they are not being polled.

 

[JohnS_3CX]

Hi @Ronin

Can you confirm if you still have issues on the final release of V16U5? (16.0.5.612)

 

[Ronin]

Hi @Ronin

Can you confirm if you still have issues on the final release of V16U5? (16.0.5.612)

Hello,

I can confirm that the polling behaviour has not changed after the update.
Still the same error:

03-06-2020 08:58:41:409 | 11 | RTL:19302.4[Extn:555] canceling call to Extn:555 with CancelSubroute Send Req CANCEL from 0.0.0.0:0 tid=c024ec42e6708 Call-ID=aQiEJyzIyKl_X-ONriTA..:
CANCEL sip:[email protected]:5063;rinstance=b2d1de712a812 SIP/2.0
Via: SIP/2.0/ ;branch=z9hGK-587-1---c0246c42e6708;rport
Max-Forwards: 70
To: <sip:[email protected]>
From: "Nel:Support"<sip:[email protected]:5060;nf=q>;tag=ab699a
Call-ID: aQiEJhYyzIyKl_X-ONriTA..
CSeq: 1 CANCEL
Reason: SIP ;cause=200 ;text="User hangup"
Content-Length: 0