Agent State bouncing between "reserved" then "ready"

[Jason Amick]

I had an isolated issue today with a single agent that would be ready and then go into reserved back ready and then reserved again.  This continued several times.  I did not see the issue first hand but after reviewing the agents logs in supervisor I can see it (screenshot attached).  I checked to make sure the agents setup was accurate as a single IPCC extension, max calls/busy trigger set to 2 and 1, IPCC line not forwarded ect.  The user has not reported any issues since the phone was reset but was wondering if anyone else had an troubleshooting suggestions?  Thanks!!

[Deepak Rawat]

Jason,

If it happened only for just one user and the issue went away with the reset of phone then I think it was probably a case of Race Condition wherein Engine allocates an agent but is never able to transfer the call successfully to that resource, worst part the thread for that resource which system should clean as a part of the process does not happen and Engine keeps sending the call to that agent back and forth causing it to go in an infinite loop of Ready/Reserved. There also used to be a defect for this in version 8 and 9 if I am not wrong, CSCue84341 if I remember correctly but you can try to search in Cisco Bug Toolkit for the exact one, not with the system right now :)

Regards

Deepak

- Rate Helpful Posts -

[Jason Amick]

Deepak:

Thanks we are running the below version of UCCX.  Agent has not reported any additional issues since the phone reboot and I have not seen any cisco bugs related for a current version. I will monitor and check with the agent tomorrow.  Thanks!!

System version: 9.0.2.11002-27

[Deepak Rawat]

Hmm, as per the defect it should be fixed in your version

 It may be happening due to some other reason as well, if it happens again will be good to open a TAC case so that they can collect logs and t/s this further

https://tools.cisco.com/bugsearch/bug/CSCue84341

Also, not sure if you are using the Dequeue step in the script to take it off from one CSQ completely and then put it in another and it gets handled by an agent there.

If that is the case, then it will be worth to apply a 1 second delay after the last Dequeue step and before the Select Resource step in the script so that Engine can get enough time to handle the events properly and does not go bizzare due to the Race condition. Only make the above change if issue happens again only else no need.

Regards

Deepak

[Jason Amick]

Thanks, if they report the issue again I will go ahead and get a TAC case opened to see if they can help.  Thanks again!!