Brian McCallion
(Guest Contributor: Brian McCallion, global performance architect at AIG.)
Cloud computing has essentially been private from the beginning. Google, for one, demonstrates that the world’s most successful search engine runs not on an IBM Power, Sun Enterprise, or other massively powerful machine engineered for business by the go-to vendors for computer solutions. More in keeping with its democratic, spare, and ubiquitous, engineering ethos Google created a “cloud” of white-box computers — throw-away hardware servers running Linux.
Rather than purchase software or hire a large consulting firm to design and build the infrastructure, Google engineered proprietary software to enable search requests and indexing tasks to be coordinated across this white cloud of white-box hardware.
Like most innovations that somehow become synonymous with brilliance and innovation, the innovation of using large numbers of identical machines to process enormous amounts of work isn’t new at all. In fact, it isn’t even an innovation that can be attributed to computing at all.
If the cloud’s origin is essentially private, then what is the public cloud?
One possible definition is that the public cloud is a disruptive channel through which to package and deliver some of the advantages of the cloud to the public.
Why is the public interested in the cloud? The cloud promises advantages that elude even the largest enterprises today. Publicly available cloud computing services are purchased by individual users, small- and medium-sized businesses, and the world’s largest firms.
What problems does the business community imagine the cloud might solve?
Traditional technology solutions take a lot of time and money to implement, yet business needs to move quickly and flexibly to seize new opportunities. The business systems of yesterday seldom adapt easily to new requirements. And the work to add new capabilities to existing systems is almost always significantly greater than the effort to build from scratch.
Given this dynamic, the risk of building new systems or modifying existing systems seems high. What makes the current practice even less appealing is the tendency over time for the cost of maintaining existing systems to grow. Some studies show that today maintaining current systems consumes seventy percent of a firm’s technology budget.
Given these dynamics, the (public) Cloud Computing model seems to offer an approach that mitigates some of the issues faced in businesses of all sizes.
1) Firms would like to be able to purchase focused, low cost, customizable, and flexible technology services
2) Pay for these services when and how they are consumed. For example, some Cloud Computing vendors offer a metered rate model in which the firm or individual pays for just the right amount and quality of resources required to meet demand.
3) Provision these firms as they are needed. If we need to provide a temporary call center in Asia for three months while we consolidate our data centers in the region, then the Cloud computing model offers us the ability to provision, configure, and host the software and desktops to do so. If a twenty-five-person firm decides that a customer relationship management solution seems like a good idea, the firm can provision and use that solution in the Cloud.
Beyond the rapid deployment, the capability to flexibly alter and shape technology services in the cloud infrastructure can help firms design, deploy, and “shape” technology solutions that fit their immediate needs, yet can adapt over time as the business evolves. In other words, compared to traditional technology practices, the financial model of the cloud seems attractive.
Chief information officers and business owners tend to look at the return on investment of existing and new technology spending. One of the key factors in the ROI model of investment is the length of time, or “payback” period over which the benefits of the expenditure outweigh the costs. It’s not an easy decision because, in the traditional model, the CIO purchases equipment, software, services, training up-front, and then hopes that the benefits can be clearly demonstrated. Yet most firms have difficulty tracking costs and benefits in a way that makes the outcome clear. If the CIO chooses too little hardware or implements a solution which the business uses later reject, the whole solution can require additional customization or additional hardware. The Cloud computing model assists in mitigating these risks by enabling both the cost and the benefit flows to be aligned. Because the building blocks of a cloud solution are much more scalable, many aspects of the solution can be tuned.
Most firms choose not to write their own desktop operating system or desktop applications. Firms make these choices every day. Yet much of the computing infrastructures today deliver little competitive advantage, they consume scarce and valuable human and capital resources.
For a firm like Google, a private cloud of white boxes orchestrated to index and return search results makes sense.
Yet, for the majority of businesses, the public cloud computing model may enable businesses to better align the cost of computing with the business value, and competitive advantages achievable through technology.
Need a Reprint?
Leave a Reply