To demonstrate the issue, on one end of the definition spectrum, the term has been expanded by many to include just about everything on the Web, or better yet, a new word to be used in place of "Web", obsoleting the term, but with nothing more to offer than an evolution of functionality. The entire Internet in this case seems to be the Cloud. You will notice in some discussions of the Cloud where the term "Web" or "Internet" is never used. This is simply too broad to be useful.
On the other end, purist types regard only virtualized servers and storage as the Cloud, and that's where the term concludes, consisting only of virtual server images and Web-scale SQL-like mechanisms. Yet I believe that is too narrow of a definition and sells the concept and its benefits far too short.
So currently, and I reserve right to change my mind in the future here in the infancy of Cloud thinking, is that I believe and think of the Cloud as a subset of the Web that can be described in three major layers.
First, at the core or inner layer of the Cloud, is the resource virtualization layer. This is the actual Cloud infrastructure, where servers are rented by the minute and storage is rented by the byte-used-per-day. These resources can be immediately expanded as usage requirements dictate, keeping costs aligned with need. It is still an abstraction from the underlying physical hardware whatever that happens to be, the software that manages that hardware, and the tools and interfaces that allow these virtual resources to be managed. The key point is that in and of itself it is all entirely useless. In other words, it's just some servers and some persistent memory, but in a logical sense rather than physical, and theoretically infinite to suit one's needs. However, applications still have to be built on it and run within it be of value to anyone other than the Cloud infrastructure vendor. Some examples of this would be Amazon's EC2 and S3, Opsource's Cloud, Rackspace, and on and on.
Second is the Cloud component layer. These are actually the building blocks of the Cloud, consisting of the APIs, Web services, data sources, usage tracking, billing components, feeds, files, and any element that can be shared and re-used in multi-tenant environments. This is where StrikeIron lives by the way, the company I work for. These are the individual nodes from which the concept of a hyper-connected universal data highway has been envisioned, where applications operate completely out in Cyberspace using celestial CPU cycles. These components are the actual doorways to the information to which the Cloud begins to become useful. If you think of one who builds on the Cloud as an artist, then this layer would be the palette of colors used to create the masterpiece, with the first infrastructure layer serving as the canvas. It is from this brush and palette of components, all existing and accessible out on the Internet, from which the third layer of the Cloud emerges.
And finally, this third layer is the application layer of the Cloud. This is where it all comes together, or the masterpiece exists so to speak. The colors (second layer components) converge on the canvas to provide meaning and therefore usefulness. These are all of the applications that are emerging that in many situations are much better suited for the Cloud rather than behind the firewall or on the desktop where they have traditionally lived. They include CRM applications like Salesforce.com (and I put Force.com here as well - if you want to know why, send me an email to bob.brauer@strikeiron.com) or Netsuite, HR applications like Success Factors, accounting applications such as IntAcct, and even development environments including Google's AppEngine as well as the applications deployed on the AppEngine. They are both built upon and enhanced by the first two layers of the Cloud, and contain the user interfaces from which a user accesses and gains benefit from the Cloud. Because of the nature of these applications and where the value is realized, upgrades are automatic, maintenance is greatly minimized, failure is dramatically reduced, and cost ultimately is more appropriate considering one's needs. These to date have not typically been new applications, but rather the old ones in a new paradigm that is enabling information technology and its uses to evolve more quickly.
And these are the three layers that constitute, in my humble opinion, the Cloud. This is how I think of it anyway.
Note what I haven't included: email (although email applications can run from the Cloud, such as Google's Gmail), consumer applications such as buying books or plane tickets, news sites, informational Web pages, online company brochures, chatting, instant messaging, and on and on. That's just the good old Web, or Internet to me. These applications of course can be built on the Cloud and and/or use its resources, but they are not the Cloud itself. Keeping this distinction is important because it ensures that the Cloud remains a subset of the Web rather than simply a synonym, and something more tangible to focus on.
That's how I see it today. Who knows what tomorrow will bring.
Interesting article as for me. It would be great to read a bit more about that topic.
Posted by: air yeezy | May 05, 2010 at 02:27 AM