Note: You are experiencing only the raw content of this site, without the intended layout and design. Either your browser has ignored the Cascading Style Sheet (CSS) files for this site, or you are using an outdated browser which does not support Web Standards. Learn more.

Home « Blogs

EDS' Next Big Thing Blog: Read and Respond to What the EDS Fellows Say About Technology

Read and respond to what the EDS Fellows have to say about the future of technology on EDS' Next Big Thing Blog on eds.com.

The data centre of the future - part 3 (what will we do with it)

by Charlie Bess

Now that we have this versatile and powerful data centre that organizations can tap into and pay for what they need, how will that change what runs within it?

This blog has described in a number of different ways the integration and flexibility aspects of Service Oriented Architectures, providing a much more business process focused set of services for the organization.

I’ve also tried to describe the new levels of simulation and model based business, providing the deeper understanding for the enterprise of its operations. The use of business intelligence, pattern recognition and business rules will drive latency out of the organization's response to information and events. With this understanding of the organization and its client's needs, companies can personalize their offerings to their clients based upon what they can “know” about what the client (not just the market) wants. We can probably all think of examples that would make our lives easier, like:

  • our bank reminds us that when our child turns 5, planning for college should start
  • our retailer would tell us when that china pattern in our wedding registry is being discontinued
  • the corporate mail system knows how to alert me to the events based upon the context of the current
  • situation

It’s not like they don’t have a good chance of knowing these things. The information today is just locked up in isolation.

With the techniques described above, they should have the capability to take advantage of this knowledge. Whole new levels of up-sell/cross-sell and fraud detection will be possible. If tightly coupled into the operational model, this business by wire or “game of business” capability should allow the enterprise to responds effectively and consistently.

I’ve blogged about mass customization and how it has never really lived up to the hype, maybe that was because it never had the computing horsepower behind it to really personalize the experience. The data centres of the future will be able to address this.

The concept of exponential growth defined in Moore's law is well understood, but computational power is increasing in another dimension as well - there are also exponentially more computers chips created every year. The information gathered or synthesized by these will provide vast amounts of data for the data centre to turn into information.With the finer grained understanding of the purchase process using RFID techniques and other information at the edge, more information can be gathered and fed into these simulations, further improving the organizations costs and making them more competitive.

The data center of the future becomes more of an enterprise value network centre, pulling together its resources, adding value through its flexible and powerful capabilities. Its limits will be on our understanding of how to take advantage of it. Even if it may be "owned" by someone other than the enterprise that uses it.

There will be those who don't like the "big brother", information rich world that this will enable, but if done well, it should make life easier for most. How this gets addressed in the laws of across the world will be something to watch.

Published Friday, November 04, 2005 4:20 PM

Subscribe to this post's comments using RSS

Comments

# Posted by Colonel Nikolai Monday, November 07, 2005 8:09 PM

Complexity against time is the black-hole of software engineering. On any system of sufficient size, the most important problem and least scalable resource will be the allocation of programmer time and their ability to comprehend the complexity of the system. Therefore, datacenter horsepower's true constraint is the people that use it. This goes for the SOA pipedream, too, where shared services suffer from interoperability problems and the amount of code needed to be written to integrate with the integration code. No, you read that last sentence correctly. You mentioned hrair. I think there is another term here, isn't it "Thairn"?

# Posted by Charlie Bess Tuesday, November 08, 2005 1:40 PM

I've previously blogged about the possibility of simplicity being the next great fad. I do have a little problem with your black and white position about SOA. Without modeling and simulation, SOA will be too complex to be operated effectively, but most modern tools and approaches take advantage of these capabilities.
I'd agree with your premise of SOA being a pipe dream, if we were talking about it all being handcrafted code written in a 3GL. Hopefully, most architects and developers are moving beyond that point into a more "process manufacturing" view of the development process. Hand coding for enterprise level integration solutions will likely go the same way as the "village smithy". There may still be people who do that kind of thing, but they don't do it around here anymore.
I enjoy getting my hands dirty and sculpting a solution from the raw bits too, but that is insufficient for most people when your using SOA at the enterprise level. I don't view tool support as being a crutch but more of a necessity, freeing people up to think about the business problem and not the commodity infrastructure issues.

By the way, I couldn't catch the Thairn reference in any useful way!?

# Posted by Colonel Nikolai Tuesday, November 08, 2005 4:17 PM

"Tharn", another term co-opted from Watership Down by software nerds like us, is a deer-in-the-headlights panic and paralysis caused by trying to solve too many aspects of a computer programming problem at once.

My experience with SOA is that it doesn't work. It requires a culture in place that doesn't exist in most big corporations (is there an echo in here?) I'm soaking in the fallout of a three-year-long SOA initiative at a very large company as I type. The initiative took the initial form of "sprinkle lightly with web services until desired consistency is achieved" and a lot of hit-and-run architects running around putting up facades that were supposed to reduce complexity, but only succeded in making fewer endpoints of connectivity. The end result is a system with incredibly bloated messaging, requiring unbelieveable complexity of new code to work with the code that's supposed to simplify everyone's life and a suite of Web Service-based protocols that will one day collapse like a house of cards when it really starts to rain hard. And need I mention that interoperability is still a total pipe dream?

Do I blame the technology? No. I blame the people for not stepping up and saying the emporer has no clothes. We did it to ourselves. The kind of intertia inside most big corporations resemble a kind of fractuously non-comittal Supreme Soviet, where everybody sees the problem but nobody wants to rock the boat of what some exec -- who has read a few too many Gartner Group reports -- says is the land of milk and honey.

Post a New Comment

: required  
required  
optional
required  
Please only click Submit once.

Subscribe to EDS RSS Feeds

I would like to receive the EDS Newsletter