Wednesday, May 12, 2010

A Cloud, by any other name, is how you tweet

There are arguments about what’s in The Cloud or, said another way, what constitutes Cloud Computing. Some purists vociferously deny that there is such a thing as a Private Cloud. Others say the only reasonable way for large corporations to migrate to cloud computing is through the use of both external, or public clouds, and internal or private cloud services.

No matter what cloud computing sellers say there are practical limitations to moving existing business technology to cloud platforms. Most often when this is discussed the issues of establishing the target platform, testing and configuring, establishing management practices covering change management, service quality, security and access and the like are central to the discussion.

The so-called mythical man-month and the related issue of how to use scarce IT funds is also central. Cloud computing proponents rightly will tell you that cloud operating costs are somewhat lower than comparable in-house costs. Cloud providers are operating at scale and usually pricing on the margin. It’s a big cost advantage… if you don’t examine the switch-over costs.

The switch-over costs are not limited to the dollars or Euros involved in migrating applications to Das Kloud, whether public or private. The act of migration is distractive. It means the halt to the development of any new function, today, while legacy applications are migrated to a lower-cost platform tomorrow. That’s a bad thing.

Smaller companies with smaller applications portfolios may feel this pinch less than medium and large firms do. But for most, the time and opportunity costs of migrating the out-of-sight, out-of-mind legacy is too high for the benefits. It is often best to move those applications as they age to the point of replacement or major re-engineering.

New systems or major bolted on enhancements are a different matter. Provided there is a reasonably current interface available these new functions can be designed, developed and deployed in a cloud environment and linked through the web to the legacy.

New uses of IT, like social media technologies in Marketing 2.0 applications, tend to be cloud-provisioned. Instead of coordinating and helping, old-schoolers often try to block corporate access to Facebook, Twitter, YouTube and the other places where the customers are learning about and discussing your products.

Companies must be involved in those conversations. This drives marketing people to external clouds and creates gratuitous governance issues for IT. IT needs to get on board. “Resistance is futile.” Whether that cloud is Private, Public, or a hybrid, for most of us a cloud, by any other name, is how you Tweet.


  1. Migration costs are always a bitch. In the old days we could at least estimate them to find theoretical crossover points. With the addition of "social" computing though we may not have any comparables. Still, at least we know that maintaining the old system on the old platform AND adding new functionality via the cloud will probably cost more in total, right?

  2. Maintaining the legacies on old platforms is likely always more expensive than maintaining them on cloud platforms (some disputes in a sec)... the trick is getting the legacy over to the cloud.

    It requires planning, execution, and testing of applications by people who, with all the downsizing, are in high demand for other things. If the legacy ain't broke, why fix it?

    My sense is the way to move forward is to put the new in cloud environments and migrate the old as it ages or as it runs into insurmountable technical obstacles.

    There are some who think that if you add all of the "proper" or "needed" change control, security management, service level management and load control to cloud computing you end up in the same place economically. I doubt it, but I've heard the argument.

  3. As Hegel once said, "The only thing we learn from history is that we learn nothing from history."

    Three decades ago (can it be that long?), PCs were the disruptive new thing against which the IT mandarins of the day circled the wagons. They not only ceded the ability to control, or at least influence innovation by embracing it, they deprived their enterprises of the opportunity to reap its benefits until they were dragged kicking and screaming into the 1980s (or out the door). In the 90s, it was the Internet. Now it's the cloud.

    The arguments against all of these seemed prudent at the time—security concerns, higher support costs, an installed base not amenable to change—until competitors who saw the opportunities turned an option into a necessity that had to be addressed in a rush.

    This is not to say that that the cloud formations now touted will be the formations that come to dominate, but clouds are definitely in the forecast. So the choice is to circle the wagons, resist, and retire early or to be a real CIO—not just an IT manager—and identify and embrace, prudently and not uncritically of course, what the cloud can do for the enterprise.

  4. Good observations on social media in the Cloud. While I believe most organizations will migrate to the cloud, there are some concerns that need to be addressed. See Computerworld - Data center forecast: Cloudy, with a chance of trouble - Feb 8, 2010 see