Compaq Software Product Description ___________________________________________________________________ PRODUCT NAME: TeMIP Expert System Access V1.0-A for DIGITAL UNIX[R] SPD 70.40.01 DESCRIPTION TeMIP for DIGITAL UNIX[R] is a family of software products for the man- agement of telecommunications and corporate networks, including fixed wire and mobile/cellular voice and data, multivendor, multi-technology networks. TeMIP V3.2-A provides comprehensive off-the-shelf fault and trouble management functions such as Alarm Handling, Event Logging and Trouble Ticketing for telecommunications network management. TeMIP supports the International Standards Organization (ISO) manage- ment standards ISO 10164-x and 10165-x, and the NMF ensembles. TeMIP and its features are applicable in the context of the International Telephone Union-Telecommunication Standards (ITU-T) X.73x and Telecom- munication Management Network (TMN) M.3010, M3100 Recommendations. It gives network operators a global view of their networks, and enables them to activate management functions and operations from single or multiple workstations. TeMIP is built on top of the TeMIP Framework, and fully benefits from the object oriented and truly distributed software architecture. The Expert System Access V1.0-A (ESA) is part of the TeMIP program and is aimed to allow plug-in Expert System solutions, to handle the ever increasing number of alarms generated by networks. As network topology grows, there is an exponential increase in the num- ber of alarms from TeMIP. Operators are confronted with a deluge of alarms that must be correlated by an expert system in order to iden- tify the underlying problems affecting network equipment. Furthermore, December 1998 AE-RDLZB-TE expert systems are also used to automate network management tasks thus improving the quality of service and decreasing operational costs. The nature of ESA means that once access has been developed, it follows the evolution of TeMIP and thus the expert system will remain upwardly compatible with it. The ESA is used to develop a customized ESA PM that enables a specific Expert System to be "plugged" into TeMIP. Three configurations may be envisaged: o A single ESA PM can be used to plug several similar Expert Systems (same vendor system) into a TeMIP director. o Several ESA PMs can be used to plug several different Expert Sys- tems (different vendor systems) into a TeMIP management network. o An ESA PM can be used to plug an Expert System directly into a TeMIP director without using a communication layer. ESA COMPONENTS An ESA PM consists of two main components: o ESA Service Provider o ESA Service Adapter. The ESA is used to customize the Service Adapter, whereas the Service Provider does not require customization. ESA Service Provider The Service Provider component forms a layer that makes the follow- ing high level services available to the Expert System by means of the Service Adapter: 2 o Collection of TeMIP Notification Events (alarms) and Configuration Events (events). o Monitoring of TeMIP Operation Contexts (OCs) to retrieve Alarm Ob- jects. This is similar to the OC Monitoring that is implemented in TeMIP Alarm Handling. o Performing TeMIP Call Requests and receiving the corresponding Re- sponses (synchronously or asynchronously). o Performing Population Requests to populate an Expert System Knowl- edge Base with the attributes of entities in a specific domain. The Service Provider supplies services to the Service Adapter using a C++ API. The services fall into two categories: o Request Services o Clutch Services. The arguments for Request Services requests are formatted the same as TeMIP FCL PM commands. They comprise a library of C++ methods for post- ing requests corresponding to: o Alarm or Event Collection o OC Monitoring o Call Request o Population Request o Expert System Failure Detection o Spawn Request. Clutch Services comprise a C++ thread, which acts as a clutch between the information read from a communication layer (if available) and the Request Services. 3 ESA Service Adapter The Service Adapter is the customizable component of the ESA PM. It is the link between the Expert System and the services provided by the Service Provider. To customize the Service Adapter, some of the supplied C++ methods are overridden (modified) to adapt the Service Provider services for the communication layer associated with the Expert System. There are three families of overridable methods that can be modified to develop an ESA PM. ESA Object Service Overridables The Object Service Overridables are a library of C++ classes, which include: o Alarm (OSI Alarm Report and OSI Security Alarm Report) used for alarm collection. o Event (OSI Event Report) used for event collection. o Non OSI Event, used for non-OSI event collection. o Alarm Object, used for Alarm Object Creation, State Change, Attribute Value Change and AO Clearance during OC Monitoring. o Reply, used for call or knowledge base population requests. o End Of Population, indicates that a population request has termi- nated. o Alarm Object Change, indicates that a given alarm object collected by an OC has gone out of the specified scope of interest, or has been cleared or modified. o Cold Restart, indicates that the ESA PM is restarting. o Watchdog, support Expert Systems that need to be polled for tem- poral reasoning. 4 The C++ classes are automatically instantiated by the ESA Service Provider in cases of successful collection, monitoring, call request or pop- ulation. To forward the information present in the objects to the com- munication layer and hence to the Expert System, the OnTranslate and OnTransmit methods of the relevant classes must be overridden: o OnTranslate is overridden to determine how the information present in the ESA object is to be translated into the appropriate commu- nication layer data structures. o OnTransmit is overridden to determine how to send the translated information through the communication layer. This overridable also allows intelligent filtering and forwarding functions to be added to the ESA PM. This overridable is only activated for asynchronous requests. Clutch Service Overridables If the communication layer does not provide a callback mechanism, it may be necessary to override the OnListenToCommunications method of the 'clutch' thread to implement the core processing of the 'clutch' services, which repeatedly executes the following steps: 1. Read information synchronously from the communication layer (or the Expert System). 2. Subsequently activate the corresponding ESA Request Services. The clutch approach is very flexible. Should the information received from the communication layer correspond to a service which is not sup- ported by the ESA Request Services, the TAL, Visual TeMIP or any other API can be used to implement the necessary processing. 5 Communication Service Overridables Two methods can be overridden to establish or release a connection with the communication layer: o OnConnect, activated by the ESA Service Provider when starting the ESA PM. This method is overridden to establish the connection with the communication layer and can also be used to start the Clutch thread. o OnDisconnect, activated by the ESA Service Provider when stopping the ESA PM. This method is overridden to release the connection with the communication layer. COMMUNICATION LAYER If a communication layer is present, the plug-in with ESA involves two threads: o Request Services are posted in the clutch thread o Replies to these services are posted in another thread. This implies that the communication layer must be thread safe or must be made thread safe using a mutex mechanism. DEVELOPING AN ESA PM The ESA consists of three subsets: o ESA Runtime System subset, a shared library (.so file) that forms the ESA Service Provider. o ESA Developer's toolkit subset with its associated include files (.hxx), which comprises two components: 1. A set of stubs (.hxx/.cxx files), which contain the default over- ridable methods that will become the ESA Service Adapter. 2. A Makefile to construct the customized ESA PM from the above files. 6 o ESA HTML Documentation subset. FAILURE MANAGEMENT MECHANISM Communication Failure Some Communication Layers provide communication failure detection mech- anisms. In such cases, it is possible to inform the ESA PM of a fail- ure and send information to the Expert System by activating the Trans- missionFailureDetected Request Service. A self-management counter attribute in the ESA PM (Failure Count) is incremented every time data cannot be delivered to an Expert System. Expert System Failure If the communication layer is able to detect a failure of the Expert System, the ESA PM can be customized to reflect this. This is achieved by calling the IEFailureDetected request service in the 'clutch' thread. There are two possible cases of Expert System failure: o Non Hot Stand-by Failure, there is no Backup Expert System to re- place the failed Primary Expert System. o Hot Stand-by Failure, there is a Backup Expert System to replace the failed Primary Expert System. Note that it is the developer's responsibility to detect if there is a backup system for the Expert System when informing the ESA Service Provider of an Expert System failure. A Non Hot Stand-by Failure results in the following: o The Operational State and Availability Status of the correspond- ing INFERENCE_ENGINE Self-Manage- ment entity are updated. o A Quality Of Service Alarm is sent with: - Managed Object: MCC 0 IEG_GATEWAY_PM INFERENCE_ENGINE object 7 - Severity: Major - Probable Cause: ApplicationSubsystemFailure o All the pending collections, OC monitoring and call requests ini- tiated by the failed Expert System are stopped. A Hot Stand-by Failure results in the following: o The Operational State and Availability Status of the correspond- ing INFERENCE_ENGINE Self-Manage- ment entity are updated. o A Quality Of Service Alarm is sent with: - Managed Object: MCC 0 IEG_GATEWAY_PM INFERENCE_ENGINE object - Severity: Minor - Probable Cause: ApplicationSubsystemFailure o All the call requests initiated by the failed Expert System are stopped. The collection requests and the OC monitoring requests are not stopped, so that the collected alarms and events can be received by the re- maining back-up system. In both cases, when the Expert System re-starts, its initialization (Collections and Knowledge Base population) may have to be repeated. The consistency between TeMIP and Expert System information is main- tained due to: o The Population Request that retrieves the current attribute val- ues of the TeMIP entities. o The OC Monitoring Request performs according to the specified scope of interest, which allows access to alarms retrospectively (Col- lection Request does not offer this feature). It is the developer's responsibility: o To specify in the Expert System Failure Detection Request, whether the detected failure is Hot Stand-By or not. 8 o Not to repeat the Knowledge Base population, OC Monitoring and Col- lections in the case of a Hot Stand-By Failure. ESA PM Failure TeMIP's distribution features allow the use of several ESA PMs in a distributed environment, but note that there can only be one ESA PM per TeMIP director. It is important, when an ESA PM re-starts after a failure, that the attached Expert Systems recognize that there was a failure, so that they are able to maintain the consistency of their Knowledge Bases. For this reason, when an ESA PM starts-up, it creates a Cold Restart ESA object. It is the responsibility of every ESA PM to override the OnTranslate and OnTransmit methods of the Cold Restart class to in- form the previously connected Expert Systems of the re-start. The Cold Restart object contains a list of the Expert Systems that the ESA PM was connected to before failure. DOCUMENTATION The ESA documentation suite provides information to install and as- sist in the development or customization of ESA Presentation Modules. It includes the following documents: o Compaq TeMIP ESA Installation Guide, AA-RDM1A-TE o Compaq TeMIP ESA Reference Guide, AA-RDM2A-TE o Compaq TeMIP ESA Development Guide, AA-RDM3A-TE. 9 HARDWARE REQUIREMENTS Alpha AXP Processors Supported: AlphaServer 8200 AlphaServer 8400 Compaq AlphaServer GS60, GS140 AlphaServer 2000 AlphaServer 2100 AlphaServer 4000 AlphaServer 4100 AlphaStation 600 DIGITAL Personal Workstation 433 au DIGITAL Personal Workstation 500 au DIGITAL Ultimate Workstation 533 au AlphaServer 800 AlphaServer 1000A AlphaServer 1200 10 Disk Space Requirements: Disk space required for installation: 14,000 Kbytes Disk space required for use (permanent): 14,000 Kbytes These counts refer to the disk space required on the system disk. The sizes are approximate; actual sizes may vary depending on the user's system environment, configuration, and software options. Memory Requirements: For runtime systems the minimum memory supported is 128 Mbytes, this also takes into consideration the memory requirements of the TeMIP Frame- work. For development systems the minimum memory supported is 256 Mbytes, this also takes into consideration the memory requirements of the TeMIP Framework. Note that if more memory is made available for use with ESA software, performance will be improved. SOFTWARE REQUIREMENTS For run-time systems: o DIGITAL UNIX Operating System V4.0D o TeMIP Framework V3.2-A. For ESA PM development systems the following software must be installed in addition to the above: o Visual TeMIP C++ Developer Toolkit V2.0 o DEC C++ V5.7. 11 YEAR 2000 READY This product is Year 2000 Ready. The testing used to confirm the Year 2000 readiness of this product included code assessment and system tests to verify transition dates. GROWTH CONSIDERATIONS The minimum hardware/software requirements for any future version of this product may be different from the requirements for the current version. DISTRIBUTION MEDIA This product is also available as part of the UNIX Consolidated Soft- ware distribution on CD-ROM. Please refer to the ordering information for each Software Media reference. ORDERING INFORMATION TeMIP Expert System Access tier licenses: Software License: QL-64SA*-AA Software Media and Documentation: QA-64SAA-H8 Software Documentation only: QA-64SAA-GZ Software Product Services: QT-64S**-** The QA-*****-H8 part numbers include the QA-*****-GZ documentation kits, but you can order an extra copy of the documentation using the QA-*****- GZ number, if required. Note: * denotes variant fields. For additional information on avail- able services, or hardware platform tiers, refer to the appropriate price book. 12 SOFTWARE LICENSING This software is furnished under the licensing provisions of Compaq Computer Corporation's Standard Terms and Conditions. For more infor- mation about the Compaq licensing terms and policies, contact your lo- cal Compaq office. DIGITAL UNIX LICENSE MANAGEMENT This layered product supports the UNIX FLEXlm Software License Key sys- tem. A floating license scheme is used, which means that the licensed soft- ware can be used up to the limit specified in the license file. The scheme used is trust based, which means that it does not use any ma- chine specific values or count of users to rigidly enforce license com- pliance. A FLEXlm key must be obtained using the request form provided with the Cover Letter, temip-license-form.txt. For use of any prior version of TeMIP, the LMF checksum is located on the Software PAK received as de- livered upon order of the Software License. This checksum is only valid for LMF, that is, for versions prior to Version 3.2-A of TeMIP for DIG- ITAL UNIX. SOFTWARE PRODUCT SERVICES A variety of service options are available from Compaq. For more in- formation on these services or other available Network Management Ser- vices, contact your local Compaq office. SOFTWARE WARRANTY This software is provided by Compaq with a 90 day conformance warranty in accordance with the Compaq warranty terms and applicable to the li- cense purchase. 13 The above information is valid at time of release. Please contact your local Compaq office for the most up-to-date information. [R] UNIX is a registered trademark in the United States and other countries licensed exclusively through X/Open Company Ltd [TM] The Compaq Logo, DEC, DECnet, AlphaStation, AlphaServer, DECwindows, Compaq, and TeMIP are trademarks of Compaq Com- puter Corporation and its affiliated companies. [TM] FLEXlm is a registered trademark of GLOBEtrotter Software, Inc © 1998 Compaq Computer Corporation. All rights reserved. 14