Product: CallPath Server for Windows NT Nortel DMS-100 Version: 2.02 Package Name: dmsw1879.exe Size: 856618 (836Kb) Level: 1879 How to install: download the self extracting file, dmsw1879.exe, to a temporary directory. Run it from any command line and follow the prompts. This will start an InstallShield Wizard to guide you through the installation process. Fixes: Internal 7876 Return control for CDN was not supported in previous release. Return control for CDN will now put the call in CND queue to default destination. PMR 1373 When extending a call to unmonitored resource, dialed number is different with switch reported number. 1) X call agent A, 2) Agent A extend to Y 3) DMS-100 reported extend to Z. the previous release callflow as: Call_alerting (A), Call_Answer (A), Call_Routed(A, null, Y) Call_Disconnected(Y). Call_Routed( A, null, Z) the current fix will reported as: Call_alerting (A), Call_Answer (A), Call_Routed(A, null, Y) Call_Routed(A, Y, Z) Note: Same flow is true. if agent a extend Y after Y disconneted then extend Z. Internal 7928 Output of error continuously made in the error logger. The error message is "Cannot Find CallID ... xxx" The condition to exit loop will never be met under this unique situation (corrupted data). Problem fixed by purging corrupted data prior to loop entry. Internal 7973 If CompuCALL is set for "Forced Message", the call flows from the switch are reversed in some instances such that CallPath's processing results in a NULL Pointer being used to do a data base search. Resolution was to test the pointer before using it. Problem fixed by testing the pointer prior to utilization. However, the event being reported by the switch will be lost. Note: The customer discovering this error found many other problems using the Forced messaging feature and has decided not to employ this function. The code fix has not been tested because the condition cannot be replicated without the Forced messaging function turned on. It is presumed that the problem has been isolated and corrected. It is strongly recommended that users do not employ this feature. If you do use this function, please advise if you find any core or trap condition. Internal 8074 DMS-100 X.25 communication code for NT did not timeout the X.25 API semaphores. Switch Connection code would wait forever if the ARTIC X.25 code for NT fails to respond to our requests. Problem fixed by timing out all "To Switch" communication semaphores. If a timeout occurs, the Switch Connection code will log the error and take down the switch connection. The error log entry contains a cause code of "semaphore timeout". Note: a semaphore timeout error most likely indicates that something is wrong the ARTIC X.25 feature on NT or the X.25 link. Internal 8086 Switch Connection code cores when it receives from the switch a DV_Call_Offered_U event with a CallMode of CallWaited on a DN that is not being monitored. This extraneous event causes the code to core. Problem fixed by processing the event without causing a null pointer exception. Internal 8142 Switch Connection code cores when an uninitialized local variable causes an address out of range condition if one of its subfunctions cannot locate a Call ID in the call state table. Problem fixed by initializing the local variable to zero. Internal 8143 Switch Connection code cores when local variables cause an address out of range condition in low memory if one of its subfunctions tries to copy memory contents based on the sizes passed into it. Problem fixed by initializing the local variables to default values. Internal 8146 Switch Connection code cores when DMS-100 switch sends it an event message with a zero length ACD DN. The DMS-100 CompuCALL Interface Specification states that the ACD DN length will be 10 bytes. Problem fixed by treating a zero length ACD DN as if the ACD DN contains ten 0's (0000000000). Internal 8147 Switch Connection code cores when DMS-100 switch sends out-of-sequence DV-Call-Queued-U messages from parallel queues. A call queued with reason of overflow on one queue is reported ahead of the call queued with reason of transfer on another queue. Problem fixed by treating calls from each queue as separate parties in one call. APAR IC27004 Switch Connection code would sometimes process TCP/IP messages received from the DMS-100 before entire message arrived. When a message was split across multiple packets, receive processing would not ensure that entire message had arrived before passing on the data. Later, this partial message would get decoded as defective and a decoder error would be logged. Problem fixed as TCP/IP receive processing now determines the size of the message by reading the message header first, determining the exact size of the message and then reading the rest of the message. APAR IC26955 When Application Logon fails due to configuration errors, the error log entry created by Switch Connection code would contain non unique return codes. From these error log entries, it would be difficult to determine the configuration parameter in error. Problem fixed as Switch Connection code now generates error log entries with unique return codes for each configuration parameter that the DMS-100 checks during Application Logon. It also includes the name of the parameter the switch says is in error. APAR IC27562 Switch Connection code treated a CDN Route in the Call History as an invalid route reason and logged a 5401 error. For a CDN, the DMS-100 now puts CDN Route in the Call History (instead of Redirected) when a call arrives at an ACD after having been redirected from a CDN. Problem fixed as Switch Connection code now recognizes a CDN Route in the Call History and treats it like a Redirected call. APAR IC27564 Stopping the DMS-100 Switch Connection which uses X.25 communication in Windows NT would cause a system icax25.exe. The system exception was caused by the Switch Connection code using local memory to pass a logoff request (shared memory is required for all ARTIC X.25 API requests in NT). Additionally, an intermittent system exception in switch.exe was occurring on shutdown in the heart beat function if a heart beat reply was outstanding when the shut down occurs. Problem fixed as Switch Connection code now uses shared memory for the logoff request and the heart beat function now handles shut down correctly. APAR IC27699 The STLMON table in the DMS-100 Switch Refrence does not match the Switch Connection code implementation. 1) STL_CALL_ACTION_PROVIDED_EVENT is supported but not in the table 2) STL_CALL_PARTY_REMOVED_EVENT is not supported but not in the table 3) STL_CALL_CONFERENCED_EVENT is supported 4) STL_CALL_PICKED_EVENT is supported (except for BCS33) 5) STL_FEATURE_INVOKED_EVENT is supported (BCS35 and above) 6) STL_INDICATOR_PROVIDED_EVENT is supported (SCAI07 and above) 7) The following events are supported for STL_RANK_PRIMARY only: STL_CALL_ACTION_PROVIDED_EVENT STL_CALL_TRANSFERRED_EVENT STL_FEATURE_INVOKED_EVENT Also, the event supported bit mask was incorrect and three unsupported events were not being checked in the Switch Connection code. Problem fixed as Switch Connection code now returns the correct events supported bit mask when an unsupported event is being monitored. For SCAI07 and above: 81090000 FF5DFFEE F0CC2000 00000000 For BCS35 thru SCAI06: 81090000 FF5DFFEE 30C02000 00000000 The following three unsupported events are now checked on a STLMON request and will now return a negative response if any of them are listed on the STLMON request: STL_CALL_PARTY_REMOVED_EVENT STL_DATA_COLLECTED_EVENT STL_RESOURCE_DELETED_EVENT APAR IC27810 When the DMS-100 Switch Connection code processed a blind conference from CallPath Enterprise, sometimes it would create a new call when the call entend request is immediately followed by the call conference request. Problem fixed as Switch Connection code now handles call conference requests occurring immediately after a call extend request. APAR IY11304 When the DMS-100 Switch Connection code restarts after an error, it occasionally encounters a 6104 error during restart which indicates that shared resources were not deleted when the switch connection went down. Once this occurs, all subsequent restarts of this switch instance will fail with a 6104 error until the share resources are deleted by running csaclean. Problem fixed as Switch Connection code now cleans up old shared resources during start up, if any still exist. Files in service: switchdm\readme.txt switchdm\csm.exe switchdm\csmpd.exe switchdm\prodinfo.msg switchdm\makepwd.exe switchdm\stt.dll switchdm\swchgdm.dll switchdm\switch.exe switchdm\swmsg.eus