Loading ...
Sorry, an error occurred while loading the content.

Why utility computing is not agile

Expand Messages
  • James McGovern
    I listened to a speech given by my own CIO regarding utility computing and how he thought this approach had merit for exploration by our organization. Of
    Message 1 of 1 , May 15, 2003
    • 0 Attachment
      I listened to a speech given by my own CIO regarding utility computing and
      how he thought this approach had merit for exploration by our organization.
      Of course, always thinking agile I had to think about how it would align
      with principles the industry should be held to and reached several
      conclusions.

      The reference to "utility" is stolen from your water, electric, phone, etc
      companies. If one were to study their models, they essentially meter
      resources and charge for utilization. On the surface this sound really good.
      Pay for what you use. But if you peel back the onion, you will find the
      approach flawed in many subtle ways.

      First, the economics of utilities favor cost recovery over value creation.
      When a utility incorporates a feature or a function, it seeks subsidy to
      guarantee its return on investment. A utility is a social construct that
      uses cross-subsidies to assure that everyone has access to the desire
      resource at "equitable" price.

      Utilities are successful in that they also guarantee standards (We all use
      the same plugs but not XML schemas) and encourage interoperability (unlike
      MS). The problem is that the utility model totally ignores many of the
      hard-learned lessons from events that effected it such as deregulation.

      The model fell apart when it was discovered that large corporate customers
      were subsidizing consumers and vice versa. You will also notice in this
      model that companies such as Enron also lost the ability to track the real
      costs of doing business.

      The pricing model also falls apart in that many utilities still haven't
      figured out how to charge for peak vs off-peak scenarios. Utilities in the
      vast majority of situations also are successful because they are monopolies.
      When one has a monolopy they tend to not value collaboration.

      When looking back into the cost recovery model, one will also need to
      consider that many large manufacturing companies are now moving their
      electrical needs to their own forms of electrical generation as the "cost
      recovery" model has made it more expensive for them to share than to do it
      themselves.

      Likewise, utilities over time also have little incentive in a cost recovery
      model to perform continous improvements (key agile principle). If one were
      to examine telephone companies, you would think that services such as DSL
      would be pervasive, but the financial model of utilities prevents many forms
      of progress.

      Utility computing is simply a financial model. Nothing more, nothing less
      and in this light should be discussed by the CFO if it has merit, not IT.

      CIOs have become savage in the pursuit of the answer that allows them to get
      a handle on reducing the amount of failed infrastructure projects and
      uncontrolled spending on capacity but utility computing may not be the
      silver bullet they seek.

      If you are interested in learning more about applying agile methods to
      infrastructure, please visit
      http://groups.yahoo.com/group/agileinfrastructure

      James McGovern
      Co-author of best selling book: Java Web Services Architecture
      http://www.amazon.com/exec/obidos/ASIN/1558609008/
    Your message has been successfully submitted and would be delivered to recipients shortly.