On Tue, 2010-07-20 at 13:41 -0400, Matthew J. Roth wrote: > --- On Tue, 7/20/10, Marko Vojinovic <vvmarko@xxxxxxxxx> wrote: > > > Say, since you have hands-on experience with this, would > > you mind helping me understand the idea itself? > > > > Specifically, assuming that I have my own hardware to set > > the whole thing up, what is the difference between having > > a server (possibly virtualized), and having a server "on > > the cloud"? And what is the main benefit of the latter > > over the former? > > > > If the server is running on your own hardware, it runs on > > your own hardware. How does the idea of a "cloud" come > > about then, and why is it any better than ordinary > > virtualized set of servers? > > > > Antonio Olivares wrote: > >> > >> This was exactly the purpose of my question, to > >> ultimately ask what you have asked here. Now to await > >> patiently the responses :) > > Antonio, > > Clearly, you are just playing devil's advocate since the answers to your questions are contained in a document that has already been quoted once in the replies to your initial post. In fact, that document is a mere two clicks away from the article that you referenced. > > Once again, from Red Hat's Cloud 101 Whitepaper <http://www.redhat.com/f/pdf/cloud/101_whitepaper.pdf>: > > These private clouds are often the evolution of a virtualized > infrastructure into something that’s more dynamic and automated... > > ...a private cloud that builds on a virtualized foundation will be > the strategic choice that helps IT bring on new business services > more quickly... > > In a cloud environment, resources can be rapidly and elastically > provisioned, in some cases automatically, to quickly scale out > based on pre-set policies and the demands of an application. Just > as importantly, resources can also be rapidly decreased when they > are no longer needed, avoiding the familiar situation of unused > servers sitting idle after the task they were initially purchased > for ends. > > It all really boils down to another layer of abstraction that can allow you to leverage your virtualized infrastructure to deploy services rapidly and get the most out of your hardware. I understand Richard Stallman's concerns about storing private data on a public cloud, but that is just one aspect of the technology. This is a pretty good answer, but I'd like to clarify a little more from my original explanation: We like to say that the definition of "Cloud Computing" is actually ...um foggy. Please forgive the marketing spin, but there's a lot of truth in this. Cloud Computing Journal offers 21 definitions of "Cloud Computing" - and that's on top of definitions from players like Burton Group, Gartner, Forrester (theirs is a doozey, by the way), and the NIST - who has a pretty good one. We defined Cloud in our company as well and I'm proud to say ours looks a lot like the NIST definition - we we released ours before they released theirs! :) So, we define Cloud Computing in my company as: (Cloud Computing is) a style of computing in which dynamically scalable, virtualized resources are provided as a service either within an organization or over the internet. Examples of services provided via a cloud include: - IaaS: Infrastructure as a Service - PaaS: Platform as a Service - SaaS: Software as a Service Things that tend to differentiate Cloud from even highly virtualized environments include: - Self Service (from the end user perspective) - Automated Deployment of Resources - Resources are Multitenant - Workloads are automatically distributed for maximum compute efficiency - Capable of existing across multiple data centers (even if they don't necessarily do so at the time) We then can get into something called the Hybrid cloud, which consists of private internal, private external, and public components. I'll leave that alone for now as I think you get the idea, save for that this is also a key driver behind interoperability standards. So, back to your question, Cloud is not necessarily another layer of abstraction as much as it a high degree of maturity, automation, and process that is applied to highly virtualized environments. >From there, running this on your own servers has immediate benefits, assuming you have applied the appropriate set of skills and operational maturity. Also note that not everyone needs this level of maturity to gain most of the benefits inherent with cloud infrastructure. Particularly for smaller implementations, a highly virtualized environment can be extremely beneficial and cost effective. Hope that helps! Cheers, Chris PS: Standard disclaimer still applies - These are my opinions and not necessarily those of my employer. -- ============================= "You see things as they are and ask, 'Why?' I dream things as they never were and ask, 'Why not?'" -- George Bernard Shaw -- users mailing list users@xxxxxxxxxxxxxxxxxxxxxxx To unsubscribe or change subscription options: https://admin.fedoraproject.org/mailman/listinfo/users Guidelines: http://fedoraproject.org/wiki/Mailing_list_guidelines
