Product: CallPath Server for OS/2 Seimens 9006 Version: 2.02 Package Name: r6o1879.exe Size: 680890 (664Kb) Level: 1879 How to install: download the self extracting file, r6o1879.exe, to a temporary directory. Run it from any command line and follow the prompts. If you answered "Yes" to the backup query and wish to undo the service update, issue the following commands: : cd \ -Qdr where is the drive the product is installed on, and is the full path of the backup self-extracting file. Fixes: DEFECT 5894 Symptom: Trigger (stop) request fails. Problem: Trigger (stop) request fails. Solution: Code changes have been made to send proper resource for the request. DEFECT 5895 Symptom: Oubound Calls provide additional routed messages. Problem: Oubound Calls provide additional routed messages. Solution: Code changes have been made to suppress the additional routed messages. Connected (or alerting) will be the first message after a Network Reached Message. DEFECT 5896 Symptom: Trap occurs during transfer scenarios Problem: Trap occurs during transfer scenarios Solution: Additional call state model checks have been put in place to avoid traps. DEFECT 6343 Symptom: When an agent_id is not supplied for an Invoke_Feature request, it fails. Agent Id is optional for logoff. Problem: When an agent_id is not supplied for an Invoke_Feature request, it fails. Agent Id is optional for logoff. Solution: Code changes have been made to allow an Invoke_Feature request without an agent id. DEFECT 6349 Symptom: Request Instruction event failure Problem: Request Instruction event failure Solution: A global variable was incorrectly declared local APAR IC08644 Symptom: System error messages such as "Error CallID not in used" and "CopyPartyList_NULL". Problem: System error messages such as "Error CallID not in used" and "CopyPartyList_NULL". Solution: These error messages appear in the system message log because of incorrect handling of control blocks in the program. These messages refers to a specific call in progress but does not affect other calls in the system. APAR IC08644 Symptom: SwitchServer does not restart automatically on CallBridge link error when configured to do so. Problem: SwitchServer does not restart automatically on CallBridge link error when configured to do so. Solution: The switch may restart due to an unexpected layer 2 frame reject error. When this happens, the SwitchServer is notified of the error and it terminates. APAR IC08644 Symptom: CallPath Server/2 does not restart automatically after trap. Problem: CallPath Server/2 does not restart automatically after trap. Solution: If there is a trap in the CallPath Server, the OS/2 segmentation violation panel is displayed which prevents the system from restarting without manual intervention. APAR IC08693 Symptom: Incorrect call_extension flag for CC/2 support. Problem: Incorrect call_extension flag for CC/2 support. Solution: The scenario is that Party A calls party B and Party B extends to party C. When the Call_Alerting message flows the call_extension flag was not set correctly to indicate that the call was part of an extended call. This prevents CC/2 from search for data involving the calls. APAR IC08877 Symptom: RCG not monitored after switch restart on CallPath Server/2 local SwitchServer. Problem: RCG not monitored after switch restart on CallPath Server/2 local SwitchServer. Solution: The application started monitors on Route Control Groups. When the SwitchServer restart, it did not restart the monitors on the RCGs which caused Call_Routed messages from the RCGs to not flow. APAR IC09061 Symptom: SwitchServer traps when call overflow to more than 8 queues and caller abandons call. Problem: SwitchServer traps when call overflow to more than 8 queues and caller abandons call. Solution: A call is queued to more than 8 queues and the caller hangs up from the call. The application receives the Call_Routed messages for each of the queues. The Disconnected message causes a trap and numerous CopyPartyList_NULL error messages in the system message log. PMR 16702 Symptom: Events are missing events when a hold-on-hold call recalls. Problem: It is possible for a call to have both parties holding each other (hold-on-hold). If one party then gets recall (starts to alert because another call they were on hung up), CallPath did not send any events. Solution: Code changes have been made to handle this case and send appropriate events. DEFECT 7864 Symptom: Originators phone doesn't autoanswer even though the Call profile as been set. Problem: An application can use the STLSAA (Set_Automatic_Answer) API to set the automatic_answer characteristic of a call profile to STL_ANSWER. This will cause party_one of a Make_call request to go off-hook automatically. A default was being assumed that is correct for CallBridge/CSA 3.1 and beyond, but not true for earlier versions of CallBridge/CSA. Solution: Code changes have been made to handle this case. DEFECT 7945 Symptom: The connect_reason is STL_CONNECT_NOT_SPECIFIED but should be STL_RECONNECT. Problem: When a connection moves from Held to Connect (i.e. is retrieved), a Call_Connected event flows, and should have a connect_reason of STL_RECONNECT. Solution: Code changes have been made to handle this case. DEFECT 7909 Symptom: Insufficient information in events to correctly track silent monitor calls Problem: When a supervisor silent monitors a connected agent, the Call_Connected event should have a connect_reason of STL_ADD_PARTY_RCV_ONLY. If the Agent is idle at the time the silent monitor is started, an extra Call_Connected event (with a connect_reason of STL_ADD_PARTY_RCV_ONLY) containing only the supervisor should be generated - it should flow after the supervisors' Setup event, and before the Held event. In both cases, when the Supervisor clears the silent monitor call, a Disconnected event should be generated for the agents' Connection_ID. Solution: Code changes have been made to handle these cases. PMR 16513 Symptom: Corrupt Program Data on incoming calls and trace/server crash application as ProgramData. This data is optional but if the data is received with zero length (unusual but legal), 2 problems occur: (1) The Call will get 512 bytes of random (probably all hex zeros) program-data (2) If a switch trace is being displayed, either in realtime or later, the attempt to format the data will fail and crash the server. Solution: Code changes have been made to handle this case. PMR 49671,999 PMR 66270 Files in service: readme.txt swmsg.eus stt.dll switch.exe prodinfo.msg syslevel.dat erarcut2.dll csm.exe engdebug.exe erarcsem.dll csmpd.exe swchgr6.dll swtr6dat.001 swtr6dat.002 swtr6dat.003 swtr6dat.004