For every vendor presentation claiming that cloud computing is an unprecedented revolution, there is an analyst paper denouncing the hyperbole. Every new example of IT goes through this phase, just as surely as a child goes through the “terrible twos” — and at about the same age.

When a concept does not catch on, the providers and analyst firms have a ready solution: rename and restart. The move to “pay per view” use of externally hosted IT resources was given several other names over the past decade, and a new bout of fever erupted each time; two examples are “utility computing” coined by Sun and HP (2000) and “on-demand computing” by IBM (2002).

No one denounced the cloud computing hype as colorfully as Oracle’s Larry Ellison during a financial analyst conference in September 2009: “I have no idea what anyone is talking about. What is it? It’s complete gibberish. It’s insane. When is this idiocy going to stop?” (See reference note at end of post.)

The following week, Oracle announced that its solutions were “cloud-ready,” which caused much rejoicing among the victims of Mr. Ellison’s rant. In fact, he was making two valid points:

  1. The cloud is not a new form of infrastructure per se; it is still comprises computers, storage, applications, and a network.
  2. As a delivery and charging model, it is not entirely new. As he pointed out, Salesforce.com had been delivering software as a service (SaaS) for customer relationship management (CRM) for 10 years already.

But at the same time, he was overlooking the significant impact that the cloud may have:

  • It changes the way computing resources are procured. As a result, the evaluation criteria and the service-level agreements (SLAs) will be different. Your supply chain department may need to be retrained.
  • It challenges the role of IT as the gatekeeper and its frequent lack of agility. If IT tells a business stakeholder that it will take three months to look into his requirements, procure an application, install it, test it, and deploy it, the business manager may have the authority (and the guts) to buy the first month of the resource he needs on the provider’s Web storefront with his credit card. IT will need to be much more responsive to remain relevant.
  • It raises significant issues in organizations that have a master data management (MDM) program that touches some of the systems used in the cloud. If your master data includes your customer list, and your CRM system is Salesforce.com, you need to be able to update the customer table in Salesforce quickly every time it changes.

Therefore, cloud computing affects the role of the enterprise architect — because of MDM, because system integration may be more complex and may benefit from service-oriented architecture (SOA), and because an application or system located in the cloud now needs to be configured rather than customized.

It requires increased reliability, scalability, and performance of the enterprise’s connections to the Internet.

So What Should You Do?

Cloud computing will probably help you evolve in a useful direction — more flexibility, less fixed costs, more focus on the business processes instead of managing the hardware and the software licenses — as long as you adhere to some basic principles:

  • Think of applying the general utility model to more than servers, storage, and applications — look around for additional opportunities.
  • When vendors get to the “What is cloud computing?” slide, tell them to skip ahead.
  • Don’t panic — it is an opportunity, not a revolution.
  • Treat security seriously regardless of the model you use. Cloud security issues are somewhat different from those for on-premise systems, but rejecting the cloud model based on ambient paranoia is misguided. If, on the other hand, paranoia has reached a higher level in the organization, then step back, educate first, and proceed carefully.
  • If your IT governance or enterprise architecture are immature, that is where the greater danger lies. You may entrust your data and your applications to outsiders under the right provisions, but you must not under any circumstance abdicate your responsibility to architect and govern your overall IT landscape. If you relinquish that role, your future will not just be cloudy; it will be downright stormy.

Reference note: Fowler, Geoffrey A., and Ben Worthen. “The Internet Industry Is on a Cloud — Whatever That May Mean.” Wall Street Journal, 26 March 2009.

avatar

Claude Baudoin

Claude Baudoin is a Senior Consultant with Cutter Consortium's Business & Enterprise Architecture and Data Insight & Social BI practices. He is a proven leader and visionary in IT and knowledge management (KM) with extensive experience working in a global environment. Mr. Baudoin is passionate about quality, knowledge sharing, and providing honest and complete advice.

Discussion

  3 Responses to “The Hype and the Reality of Cloud Computing”

  1. avatar

    “If IT tells a business stakeholder that it will take three months to look into his requirements, procure an application, install it, test it, and deploy it, the business manager” will be ecstatic…..ecstatic but disbelieving if they’ve been around the block.

  2. Cloud is for Flexibility and NOT for cost saving. If you approach cloud as a means of achieving saving cost you are likely to be disappointed. However, if you approach it which a mindset of achieving flexibility, you will get better result.

    http://setandbma.wordpress.com/2011/06/15/cloud-is-for-flexibility-and-not-for-cost-saving/

  3. I would say that Cloud Computing as a concept and a business model often helps to manage many of the well-known IT project risks – starting from large COTS deployment – and save some $$$, however, it also add some new risks as well. Would be great to see more Cloud related risk management papers and industry studies.

    PS…I am trying to be polite and accurate, you see.

 Leave a Reply

(required)

(required)

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>