Cloud Computing-Not a Product or a Technology
If you believed all the media hype and marketing materials, cloud computing would be the silver bullet of IT in the 21st century. It makes companies more agile and more effective, reduces costs and optimizes the delivery and operation of IT to make them simpler, more flexible and faster. Meanwhile, back in the real world, most cloud computing projects collapse in the planning or pilot phase. Why? Generally, "shortsightedness," a lack of orientation and experience, the wrong tools and, most importantly, misguided approaches. In other words, cloud computing is more than just a product or a technology.
A Bug in the System
Cloud computing is a complex concept. Indeed, for many companies it is apparently so complex that they have written it off as something no longer of interest or value to them. They shelve their projects until some distant future, hoping other companies, manufacturers and industry experts will eventually provide insight and clarification on the subject. Other indicators show that many companies don't even get past the planning phase, choosing rather to cut their losses in the early stages of a project. In other cases, flawed assessments and methodologies when integrating their existing IT operations prevent most companies from reaching completion. Controversial questions and unexpected barriers arise that seem impossible to overcome.
In almost all of these situations, one major mistake is made: Cloud computing is treated as a purely technology-driven concept. But to achieve the established objectives of a cloud computing project, much more needs to be done than just changing the technological infrastructure of an enterprise. The IT organization, IT operations processes, the skillsets of IT staff members, and a major shift in management and employee mindsets all play significant roles in the overall project strategy.
Staying Focused
One thing in particular drives a successful cloud computing project: scalability. This relates not just to technological elements of the project, but also to any operating costs and capacities required to make it a reality. But beware of making grand conclusions. With regard to overall economic viability, it is important to remember that cloud computing does not necessarily lead to an absolute reduction in IT personnel. Instead, it tends to shift existing operating capacity to planning and management activities.
This shift is then accompanied by a mandatory adjustment in the tangible and intangible skills of a company's IT staff. Although expertise with regard to the latest technologies and products are still a dominant force here, IT service management competence needs to grow.
IT organizations have changed dramatically in recent years, from primarily decentralized, product-oriented units to more centralized, technology area-focused entities. In fact, many companies have made attempts to establish their IT as "in-house shared service providers." In reality, however, not many have actually achieved that goal because they are being confronted with, and stumbling over, the same obstacles that are once again popping up in the cloud computing arena: The "Taylorist" division of labor in IT organizations (based on Frederick Winslow Taylor's theories), a breakdown into "silos" or specialist departments.
Cloud computing requires a horizontal, service-oriented structure, which in turn necessitates an interdisciplinary team approach and a technology- and product-based system of accountability—an end-to-end treatment of all existing service elements. To do that, structural changes have to take place in an IT organization; a shift must be made to an integrative model that more closely links the vertical functions of the IT value chain.
The operational model of cloud computing can be compared to a modular assembly line in a factory. The main challenge of the model is to create a highly available and elastically scalable platform on which standardized IT services can be set up in a modular manner that suits the customer in question. To guarantee this, focus must be placed on planning aspects such as capacity, availability and service management. In addition, operative processes such as change and release management need to be more consistently assessed and more effectively implemented. Contingent on the requirement of reaching higher automation levels, implementing provisioning concepts and achieving improved standards of integration among the various virtualization technologies, the operating model also changes with regard to activities and tasks carried out by IT administration teams. Running, maintaining and supporting the platform becomes a deciding factor in quality here.
Much has already been written and discussed about the invisible wall that apparently divides IT and end users. For cloud computing to be successfully implemented, this barrier must be eliminated. One important aspect in this is how IT departments and end users see one another. End users tend to see IT as a gatekeeper—users want fast, easy and flexible access to required IT resources, data and information, but in their minds, IT limits or prevents access. One of the negative consequences of this mindset is currently being publicly debated under the topic of "shadow IT." Users are trying to purposely avoid IT through their own independent consumption of public cloud services.
In contrast, IT often sees its end users as opportunists who, despite looking for efficient business solutions, are not required to run those systems and therefore do not understand or appreciate the difficulties and complexities of IT's tasks.
IT needs to reinvent itself a bit and present new solutions that make the "provision" and use of IT services faster, easier and more flexible in order to generate among end users an increased appreciation for the concepts and requirements of IT. A cloud computing strategy is a good approach here.
Old Technology, New Environments
The underlying technological concept for cloud computing consists of individual solutions that companies have used for many years, but need to be newly tailored and sensibly enhanced for their new purpose. Virtualization technologies and terminal services, for example, have been in use at many companies since the 1990s, while automation and provisioning are known quantities if we take into account software distribution processes or the mechanisms of some Web hosting companies.
At this point, even the integrated environment in which these technologies are merged has been available for many years. Just think of application service providers (ASPs) who have been using this platform for ages. What is new here is that the concept is being used as an enterprise approach. In order to best support that effort, software and hardware are being increasingly customized for the job.
A quick search on Google for "cloud computing" or "cloud solutions" will provide you with a bafflingly long list of "cloudified" software products—they are popping up like weeds from the ground, or like cumulus clouds in the sky. All of these software solutions are equipped for cloud computing, include the buzzword in their product names, or belong to a whole "family of cloud software products."