Mentally I have a multi-tenancy problem. On one shoulder I have a hippie wannabe creative type that bangs on musical instruments, writes, paints, and focuses on the creative process. On the other side, I have a little suit-wearing business analyst. This duality is sometimes maddening. The artist in me wants to poke a stick in my temple when I hear “the cloud” on yet another TV commercial, and the other part of me is excited to be part of the next phase of this technology cycle. Now that I have everyone questioning my mental state… let me provide a diversion and state that duality, diversity, alternating needs are part of what is driving the Cloud movement. So let’s talk about what’s really happening and why everyone is forced to watch those crazy commercials.
“Value of the Cloud Today”
As I describe the value of “Cloud” to companies today, I don’t want to claim authorship on my categorizations. Many have provided inspiration. In an effort to recognize all, I’ll point out Dave Vellante’s iterative work on “Stack Wars” as a good example. He makes many good points about the race to technical dominance among the big vendors today. With that said, I’ll continue on like it’s all my idea. For my discussion purposes, I’ll overly simplify and say we have 4 operative categories of computing today:
1 )Traditional Tech: For this discussion, I am referring to everything from 2009 back, Client Server, Mainframe, etc. Over the years since we first used Hollerith’s Desk to tally the census data around the turn of the 20th century, we have worked to optimize individually owned computing environments and we have spent trillions of dollars on our data centers in primarily a self ownership model. For more info on cost, this blog has some of the details on IT spending ( Information Technology Spending By Country, etherfire, brighthub.com, Published Jul 31, 2010). Because of its dominance, we will not see traditional tech become extinct any time soon, but I do see notable changes in the following 3 categories.
2) Vertical Integration: Henry Ford was a master at Vertical Integration. He owned the rubber tree farm to make his tires, machined his own parts, etc. This is a very traditional model for the beginning of a technology curve (aka automobiles in early 1900’s). Interesting enough some companies are deploying this strategy today in the later, more mature phases in which we are currently. A good example is Oracle’s exa-N product lines. It’s an “all in one box”, single vendor strategy. The focus is specialization and is aimed to increase performance for data analytics and potentially reduce costs from server reduction. There are other examples of “IT appliances” in the market. The company where I work has similar purpose-built appliances. These purpose-built “plug in and go” appliances will continue to gain ground because they are easy to consume, but they need to balance their specialization with their cloud integration capabilities. Why should you care if they connect to the cloud? Because, history proves specialized/isolated assets are more expensive than communal assets. Appliances will likely be “cloud compliant” or they will find themselves as a tangent away from a cloud strategy and obscured from the primary plans of the business.
3) Virtual Integration: This technique leverages the features of Virtualization to blur the lines of proprietary systems into an abstracted layer of logic technology components. With a virtually integrated environment come optimizations for utilization, improved flexibility, notable savings from consolidation and other benefits. Every virtualized environment is not necessarily a “cloud”, but it provides a medium to begin developing systems with cloud attributes. This still requires the individual companies to retain ownership or to bundle managed services as part of their efforts.
4) Hosted Service Levels and/or Functionality: Hosted environments provide a utility type computing environment where the user can consume IT units without a great deal of day-to-day knowledge about what is happening “behind the curtain”. Hosted environments are not the sole domain of the Cloud, but they provide a scaled platform and a financial transactional model to monetize or apply the cloud to IT. In the last decade hosted systems have grown tremendously, albeit not without many customers returning to in-sourced or on-premise systems. These reversals are a mix of growing pains in scaled execution and changing customer requirements. Though I’ll point out change is a function of clouds. Clouds should allow for IT units to “float” between providers.
Within some of these categories you can begin to see optimizations achieved by the abstraction of the physical equipment from the logical functionality and the utilization of service that abstracts the consumer from the technical expertise. Both approaches are optimizing through a function of consolidation. These changes mirror the evolution of electricity production (expertly outlined in “The Big Switch” by Nicholas Carr, a book and topic I plan to cover in a later blog.) These building blocks provide the technical stage for a new business model we call the cloud. Today we have 2 cloud formats “Public Cloud” and “Private Cloud”. A Public Cloud is where a consumer buys the service and uses it paying per use, a monthly or an annual fee. Public Clouds generally imply multi-tenancy. Meaning my stuff is next to your stuff and there is some basic security to protect us from each other.
A Private Cloud provides cloud-like functionality, but is housed in a single instance either owned by the consumer or specifically hosted for that consumer by a 3rd party. Whether a consumer leverages a cloud publicly or in private, there are some basic benefits they are able to leverage:
• Utilization – Because your servers for say Exchange 2010 are now virtual, you don’t have to purchase all of them as physical assets, they will co-exist on a large physical server, and this inherently increases utilization as an example.
• On-Demand Capabilities
o Performance – Yes VMs do add a bit of overhead, but once again you’re combining strength. Imagine 100 cars on the highway and every car going downhill gives their horsepower to the cars going uphill. This will increase the ratio “performance/dollars spent”.
o Capacity – Similarly, capacity needs fluctuate and virtualization can help level set to an overall optimization of available assets.
• Virtual Management– Today Cloud technologies take the fundamentals of virtualization and add management capabilities which are critical to enable the switch to the cloud.
o Separation of duties – ability to define delineation between users and functions. Strong permissions controls, performance controls of the virtual machines, etc.
o Simplification – Example, if you had a box crash 100 miles away, you would have to requisition a new server, wait for it to ship, plug it in, and load it with new software. If that box was a VM, you could send the server configuration like an email to the remote server and enable it with a couple of button clicks. Cloud simplifies, cloud tools provide templates and wizards to increase the utility of the administrator.
• Owner to Consumer Switch – With all this new flexibility and simplification, companies can evaluate and execute a more fluid ownership model. Do they own this asset or do they ship it to a partner? Do they allow an integrator to host a developing system and move it back when it’s ready for production, do they own the primary system, but bolt on clouds from 3rd parties? These are options that were non-existent or at least much more difficult to execute just a couple of years ago.
“What’s next?”
Going forward the Cloud will provide greater utilization and flexibility in how we consume IT units. In my discussion, some would say I missed a Cloud category when I mentioned only public and private clouds. That would be “Hybrid Clouds”. I did this because to me this is how the cloud will be applied. “Hybrid” is a mash-up based on the customers financial and risk requirements. For me, the application of cloud will be like a Cloud shopping mall where you have a diverse series of store fronts offering a variety of ownership models and service levels. Figuratively, you may choose to buy a pre-made suit for one need, and the thread and cloth for another. I must assume this will be good for the customer and our collective GDP. My Business Analyst personality likes the options for business going forward, and my hippie side likes the fluffy clouds painted on everyone’s logos these days.
Andy Sitison 