Questions of element management
Element management is rarely addressed - perhaps because on the surface its role seems fairly dull, or perhaps because it has the potential to annoy just about everybody.
But at the dawn of an age where management systems are beginning to automate multiple processes (fault, performance, service, etc.) across multiple infrastructure components, we have to ask what role element management plays in all of this.
An element includes network devices and systems of all types, along with their required firmware, operating systems, etc. At the highest level of abstraction another element is software - including individual applications and database management systems.
Element management should address fault, configuration, performance and security of individual network elements. But this doesnt usually happen in a vacuum - so lets say, " The role of element management is to provide appropriate information for fault, configuration, accounting, performance and security (FCAPS) management, as integrated with the relevant management software. "
You may be asking what the relevant management software is. This question has been haunting the industry for years, and the answer will always be changing.
Another, far less often asked question - one that may hold the key to redefining element management itself - is, " What is the APPROPRIATE information for integrated management, and how should it be shared? " In other words, between the network device community and the enterprise and OSS management software community, who should do what, under what circumstances, with whom, to set the stage for integrated, efficient, and more fully automated management? As far as I know, there are no yardsticks for this. Beyond looking at MIB implementations, its a question most people dont want to ask.
Why is that? I would suggest that it carries within it the potential to offend just about everyone. To answer it, network device manufacturers must abandon once and for all the idea that management is an add-on enhancement designed primarily to get people to buy their hardware. They will have to view management as a separate, multibrand universe, in which they will have to find a creative way to add value while remaining good citizens. And in so doing, it opens up a world of complexity that, as a class, they would rather avoid. For the broader management software community, the independent service providers (ISV), it punctures their comfort level - demanding products that can stretch across many different elements. As such, it opens up a world of complexity that, as a class, they would rather avoid.
EMA is planning to research this troubled area across the full set of FCAPS disciplines, within both enterprise and OSS environments. For this I would like to solicit your questions. With element management, which vendors would you like to see change and why? Are you faced with a set of element and process bigots who dont see eye to eye? Are you struggling to cobble together performance, fault, quality-of-service and accounting solutions among those offered from device vendors and those provided by ISVs? If you could recreate this universe from scratch with a new set of rules, what would those rules be? I look forward to hearing your ideas.