Next-gen architecture: fantasy or reality?

Why bother to even think about a next-generation architecture? Is it something best understood through standards, such as those behind Web services, or are standards sometimes actually inhibitors? More importantly, what’s a next-generation architecture for – what good does it do? Is it even real, or is it just science fiction in analysts’ clothing?

As I mentioned last month, Enterprise Management Associates is developing a paper on next-generation management architecture – where we see it going (or perhaps more specifically, where we hope it goes) and why. So if you have any thoughts you want to share, please e-mail me at mailto:drogseth@emausa.com

I’ll be the first to admit that there is an element of imagination that must take hold if any industry progress is to be achieved. Leaps in science are rarely purely deductive; they are a little more like art. Through dialog, experimentation, and the natural business pressures that favor certain design points – some of these “flights of fancy” emerge victorious over traditional ways of doing things. Most of the imaginative quests die on the vine.

Addressing the role of standards is important. Often, industry evolution gets “hyped” around certain standards, or protocols – from OSI to SNMP to Java and all its affiliates, to CIM, XML and SOAP. Standards can be fundamental, or they can occasionally lead to “jousting at windmills” on a global basis. Standards, particularly those that would ease integration and communication between geographically distributed management products, are must-haves for a next-generation architecture to emerge. On the other hand, I would state just as emphatically that no standards can truly capture what a next-gen structure should look like. Standards are not intended to address that. They are just technical enablers.

This leads to the final – and most profound and difficult question – what’s a next-gen architecture for? If you can truly answer that, then you can begin to design it. The reason that this question is difficult is it must start not with architecture at all, but instead with human behavior, organization, business dynamics, psychology, desire and so forth.

Ultimately, the industry is designing products to empower customers to succeed in work and be happy in life. In IT infrastructure management these values have recently translated into the vision of on-demand computing, which makes the following two assumptions:

* IT must do more with less.

* IT must align with the business.

As general as these assumptions are – and they don’t truly address details in the cultural evolution of IT – they seem like good departure points. EMA has had to make assumptions about the nature of business evolution, IT evolution, and cultural and organizational options prior to even considering developing an approach to next-gen architecture. Some of these have been addressed in a paper on IT evolution – some aspects of which have been summarized in past columns.

I’d like to end by explaining my own personal model of next-gen management. It has to do with CDs and CD players. (And yes, I know the Internet is rapidly transforming that model – but for the moment, never mind.) Historically, management products have been developed as separate silos. As they emerge, those involved with data gathering, data store, data transport, integrated security services, infrastructure capture and baseline analytics are just beginning to move towards what I call a “CD player model.”

To some degree my idea of the perfect next-gen framework is a CD player, over which different products, rich in their own analytics and visualization, can be played. This model is admittedly similar to Microsoft’s business model, with operating systems and applications that run over them. From a management perspective – that’s my elevator pitch for this vision – there’s maximal choice, maximum efficiency, maximum incentive for industry evolution.

How would you like to see the industry evolve? Write me at mailto:drogseth@emausa.com