Let's now unpack some ideas and put some real technology and project management flesh on them. This is needed, as to quote the old marketing slogan, we face a lot of FUD (fear, uncertainty and doubt) when it comes to Cloud or perhaps, 'fog' instead of FUD.
In practical terms, Cloud means using a lot more remote services delivered securely over IP (Internet Protocol) than the traditional model of an organization utilizing the services of an in-house IT department. Cloud-like functions have been around since the start of commercial data processing – what else was the classic payroll bureau service than remote provision of a computing resource, after all? A key difference, economically, to earlier versions of 'Cloud', even classic outsourcing or managed hosting, or even the first generation of ASP (application service provision): customers should only pay for what they specifically use.
An analogy often employed in the literature is that of rent versus buy; the former entails a low-cost short-term, flexible arrangement between vendor and customer, the latter connotes a larger investment where responsibility devolves immediately on to the purchaser.
So far so good, but then the IT industry being the IT industry, we are soon enveloped in a mist of terms ending 'as a service,' as in 'x' (Platform, Infrastructure, Software), but made more opaque with nomenclature like 'PaaS,' 'IaaS' and 'SaaS'. Let's not disturb qualify complexity needlessly, but the real point is that when choosing a cloud version of an IT function you as the customer should be able to expect to use it without having to worry any more about its internal density, a black box approach, if you will.
'Public,' 'private' and 'hybrid' Cloud
At the same time, you definitely need to know the general shape of the profiled Cloud as they differ, which has obvious consequences in terms of their security and cost to you as a subscriber. The most important concepts are 'public,' 'private' and 'hybrid,' so let's briefly define those. A public Cloud is what you get when you as a consumer use Google Apps or back up your movies and digital content in something like Apple's iCloud. Data is protected by password access and held in the data storage facilities of companies like Amazon or Microsoft (with its Windows Azure) who offer this resource to all comers.
The emphasis is on getting rid of the need for local storage (as in your desktop or LAN) and move it to a remote 'back end'. Note that public Cloud solutions are not just for the consumer, indeed far from it; many organizations use solutions like Amazon Web Services for the same ends, for example. This approach is fine for most consumer uses but also applications that are lower down on the mission-critical rating.
A private Cloud uses the same model technologically but now you as a client contract with a company (eg a cloud service provider) who offers dedicated infrastructure and facilities that are specific to your needs alone. Patently, that means a higher level of service to you, especially when it comes to security, access and protection of your information and IT services.
So, again, you move away from local data storage and data processing to a more remote model, but instead of sharing the resources with others, you have effectively put your name down on a solid chunk of Cloud functionality for your use alone. This model is providing appealing to mid-sized and larger organizations who like the flexibility of the Cloud but who dislike the idea of it being as 'shared' as the former approach.
A hybrid approach is not just a 'mix' of the two. It is in fact a bespoke environment designed to meet the different needs of business. One approach is for a Cloud provider to offer a customer a mix of dedicated remote services via its own private Cloud with others distributed via a public Cloud.
An alternative approach may be where the customer maintains a structure with the majority of services fully in-house, some on a managed service basis, and some in a private Cloud whilst provisioning services such as online data backup, email security, archiving, and disaster recovery and so on in a fully public Cloud.
The best way to understand all this: think of Cloud as a matrix with one axis marked 'closeness of data to me' and the other 'closeness of relationship I need with the third-party looking after my data'.
Source by Paul A Cook