As an alternative to the three tiered classification of cloud based services – Infrastructure as a Service (IaaS), Platform as a Service (PaaS) and Software as a Service (SaaS) – cloud computing can also be classified based upon how the underlying networks are implemented and accessed by the client. The following article provides and overview of the two principle models Public and Private Clouds.
These models do not describe strict architectures and so a variety of configurations can be employed to achieve each model of cloud computing. Furthermore, each model can be used to offer each tier of cloud service as mentioned above.
Public Cloud
When most people think of cloud computing they think of a typical public cloud model where the services on offer are available to public customers through a public network, usually the internet. The accessibility of this model allows for cloud services to be supplied to a mass market and for the provider to centralise/pool (and share) the physical computing resource they require to offer the services; utilising virtualisation where they need to establish operating environments. As a result, consumers and providers of public cloud services can benefit from considerable economies of scale to reduce their costs as well as sheer scalability to respond to fluctuations in demand.
Public cloud services can often be free to the consumer, funded for example by advertising, however where a charge is applied they usually follow a pay as you go (PAYG) type model – that is the consumer is supplied with the computing resource they need as and when they need it and are only charged for what they use. Many of the cloud services that are most familiar to the general consumer, particularly Software as a Service (SaaS) applications such as Google Docs (cloud based ‘office’ applications), Gmail (web based email) or Dropbox (cloud storage), fall under the public cloud model.
For some consumers, the mechanisms of the public cloud which bring such cost savings and widespread availability can also bring undesired security vulnerabilities. Although the security of these services is always improving, the fact that data is transferred across public networks and stored on shared physical and/or virtual servers may provide obstacles to some clients using the public cloud for processing sensitive data. Instead they might look towards to private cloud services.
Private Cloud
Cloud clients who are handling, processing or storing sensitive data that they need ensure remains private and secure and who want to take advantage of some of the cost, availability and scalability benefits that cloud computing usually offers, can utilise private cloud services as an alternative to the public cloud.
The concept of a private cloud is harder to define than that of public cloud and no strict definition really exists. It is best to categorise private clouds by some of the features that they provide and the issues that they address in comparison to public services, such as data security and ownership control (of servers etc), rather than the mechanisms they use. They will of course also display the tell-tale features of cloud computing in general – virtualisation, instantaneous scalability, PAYG charging and automated allocation of resources on demand.
The entirety of a private cloud will only be accessible by a single organisation (in contrast to multiple clients using services within the public cloud) and some examples therefore also allow the end user organisation to have control over the management of the cloud network and its set-up so that they can tailor it to their own needs. The flip side of this control however is that some of the cost savings generated by the centralised management of public cloud services will be lost in private clouds. Instead the service becomes closer in structure to a traditional local network based service, but with the benefits of virtualisation mentioned above.
The network of servers itself can be hosted internally by the organisation (as required by some regulatory bodies for sensitive data) or externally by a cloud provider (e.g., a Virtual Data Centre, VDC ) but access will be restricted to connections made behind the organisation’s firewall. To provide the security and control private clouds can employ certain characteristics techniques including closed networks of servers (not shared between clients), hosting of servers on site and leased lines to access these networks where they are hosted off site.
Recommended Posts
Interpretivist social scientists, by contrast, could use social critique or symbolic interpretation somewhat than developing empirically falsifiable theories, and thus deal with science in its broader sense. In trendy academic practice, researchers are often eclectic, utilizing a number of methodologies . The […]
2012 is the year of discovery and change which will revolutionise the way we see our world and will prove the limitations of scientific theory, hinting at the wonders of the infinite and the unknown. The matter in our Universe, when it […]