After MI and softphone issues were resolved, this week we finally rolled out Release 1 of the solution for this client to the remaining Contact Centre sites. This means we are now live with approximately 2000 advisors handling 35K+ calls per day.
As usual the go-live was not quite as smooth as we would have liked. By Monday afternoon we were starting to get calls from Customers complaining that they had waited a long time in queue, hungup, called back and got answered immediately.
Realtime and historical statistics gave no indication of the problem. After a lot of digging by the team we identified that calls were getting stuck on external routing points (ERP) between the Genesys SIP side of the solution and the Avaya side of the solution.
The root cause of the problem was the Class of Restriction (COR) on the Avaya stations was not set correctly on a small number of stations. This resulted in the call setting stuck in the vector associated with the ERP (VDN). Unfortunately these advisors kept getting targeted again and again which meant that the total number of stuck calls was much higher than the number of incorrectly configured stations (for obvious reasons I won’t tell you how many calls were affected by this problem!).
As a secondary problem we found that there was no skill applied to the ERP VDNs. This caused the call to enter a “black hole” since as a backup even through the associated vector was set to target advisors there was no Avaya skill / split specified.
Further analysis showed that the maximum time to abandon was 99 minutes. We tracked this down to the “Vector Disconnect Timer” specified in the ACM system parameters: