KEY CONSIDERATIONS IN HP NONSTOP ENTERPRISE MANAGEMENT :

IT Article, 2003

One of the observed shortfalls of integrated solutions for an Enterprise Management approach is the lack of appropriate levels of security, auditing and contingency for each proprietary platform monitored. In providing an integrated, end-to-end graphical view of service provision, the rudimentary design considerations engineered into key, industry-leading solutions should also be securely extended to the users of the enterprise layer of the business. This consideration is especially highlighted when relaying critical events away from a ‘NonStop platform’, to be dealt with by non-platform specific solutions. If this is to be the chosen approach for the business then consideration should be given to a number of key areas, which are highlighted in the data flow diagram.

Maintained Security for both Logon and Manual Task Initiation
A multi-tiered approach to security at the enterprise layer is imperative when allowing remote users access to the hp NonStop platform. Authorisation to access NSK tasks for remedying problems reported to the Enterprise layer of the business should be enforced locally, or on the hp NonStop platform. The user database should be encrypted and secure. The ability to allocate event-viewing functions, task functions and other interactive features as separate user facilities at the enterprise layer is extremely beneficial. This will provide a more secure and controlled environment that allows for both junior and administrator level access to be defined. Individual tasks should be split by type, thereby allowing tasks relating to each application or system area on the hp NonStop platform, to be linked to an appropriate user responsible for that business function.

Comprehensive Auditing of all User Actions
All user actions relating to task initiation should be audited. This should include the time at which the task was started, the Enterprise user initiating the task, the platform the task was initiated from, e.g. a WIN32 GUI at the Operations Bridge, and the NSK task name that was initiated. Audit reporting, historical look-up and archiving functions should also be provided.

Supplied Troubleshooting Functions should be Maintained
The hp NonStop platform is unique in its provision of probable Cause, Effect and Recommended texts for each event generated to the primary collector. In relaying events away from the NSK platform for remote viewing, it would not be sensible to denature this troubleshooting function for those who are less knowledgeable in hp NonStop related issues. The option to allow these texts to be viewable at the Enterprise layer should be provided where the Enterprise Manager is being used for problem resolution as well as SLA monitoring.

Ensure that NonStop Mechanisms have been Employed
A key sentiment for those concerned about employing Enterprise Management platforms not traditionally associated with fault tolerance, is the exporting of NSK duties to those remote boxes. Needless to say that if this model is adopted by the business then NonStop relay mechanisms should be implemented by design and incorporated within the chosen solution. The solution should incorporate NonStop dynamic re-routing of events to a secondary box in the event of failure. It should also allow connections to the hp NonStop platform (for task initiation and status checking), to be transferred successfully in the event of partial TCP/IP process failure.