An alternative way to isolate Genesys or switch call path issue, please try to set call_answer_type_recognition to "no_progress_detection".
call_answer_type_recognition Call Progress detection mode; valid values are:
(a) no_progress_detection
(b) no_am_detection
(c) positive_am_detection
(d) full_positive_am_detection,
(e) accurate_am_detection
(f) telephony_preset
This option can be configured in OCServer properties as global value or in campaign object's properties as local campaign-level value.
If dropped call issue occurs still, it could be caused by switch call path issue.
To quickly verify voice call path (with logged-in agent, of course), make a call to the "origination DN" of campaign agent group object.
What transfer method is being configured? 2-step or single-step?
Is Genesys URS server being used? It would be 'easier' if URS is in place.
Behaviors of symposium-link (as in 1Q 2007) + Genesys T-server (7.2) are quite difference from previous version.
Once CDN is registered and defined as Route Point or ACD queue object in CME, CDN's CNTRL flag is set to YES automatically; the flag required manual configuration in switch in some previous Nortel switch/Symposium versions.
Impacts:
- CDN "cannot" be defined as ACD queue object in CME (Genesys T-server is monitor-only to the CDN and has Nortel switch to do call routing).
- If ACD queue object is defined in CME anyway, Nortel switch sets CDN's CTNL=YES and waits for 4 seconds for every call to let external application to do routing.
- Please correct me on this if I'm wrong! Any idea to configure such case (switch CDN CTNL=NO as ACD queue in Genesys monitor-only) is appreciated!