The 'cloud' is a convenient generalisation for a diverse set of products and services. Read our simple guide below.

Cloud hosting solutions vary greatly, and many ask the same question: what is cloud hosting?. Even if they appear the same ‘on the outside’, they’re not. Their underlying infrastructures, features, interfaces and layers are unique. That being said, they can organized into three general categories, each defined by the architecture of the solution. And that all-important architecture determines or influences the level of control, flexibility and transparency each delivers.

Cloud 1.0 Solutions

This type of cloud platform is considered a Software-as-a-Service (SaaS) model where a customer’s app has code running against a background server to create individual servers: web, database, compute engine, CRM, e-commerce, etc.

Essentially, the customer is running on a shared server with other customers of the hosting provider – meaning that data input, output and processing become co-mingled with other customers’ data during operations and transactions.

Advantages

• Very inexpensive (Venture Capitalists tend to endorse this cloud model so that money goes to software and apps and other corporate expenses)
• As a highly-controlled environment, they tend to support prefab plug-ins that are easy to implement, even if they are inflexible.

Disadvantages

• Limited or no flexibility in app development, since the highly-controlled environment compels customers to write towards a very specific AP. Differentiating one’s app from a competitor’s becomes difficult if not impossible
• The customer is at the mercy of the server’s internals or engine which is hidden from them. Developers and architects cannot see into the foundation of their cloud, making it difficult and inefficient to work with in terms of scaling, projecting costs, optimising the environment and securing their data. Developers have continually to write and rewrite code to optimise their environment against an ambiguous cloud environment
•  The co-mingling of  data with other customers’ poses many security and performance risks

Cloud 2.0 Solutions

This type of public cloud differs hugely from Cloud 1.0 solutions, in that virtualized instances of servers run across a massive underlying hardware environment and tap into a limitless reservoir of raw computing power.

Customers create their own virtual instances for web heads databases, security applications, etc, as required. The computing resources are dispersed and absorbed across the hosting provider’s entire cloud environment and their data inputs, outputs and processing are not co-mingled with those of other customers.

Advantages

• Better security since every customer is running independently
• Better performance as customers manage their own instances and resources
• Literally limitless scalability because the addition of new virtual servers is not limited by the rigid limitation of physicals as in Cloud 1.0 offerings
• Cloud 2.0 encourages and permits a more collaborative approach between admins, architects and developers to create a smarter, more efficient scaling environment

Disadvantages

• With greater flexibility and control, customers need to make choices about their utilisation of cloud resources to get the most out of them
• While not a real negative, many perceive Cloud 2.0 solutions as being less secure, as with Cloud 1.0 solutions. But this is an incorrect assumption, since the hardware foundation of this type of cloud typically has security embedded within it, rather than running over it.

Private Cloud Solutions

Private clouds are, quite simply, Cloud 2.0 solutions running on a customer’s dedicated hardware. And, like any cloud, they differ according to each provider’s infrastructure and hardware.

Advantages

• Customers can realise an even greater degree of flexibility and control in how they administer their environment
• There is often an opportunity for customers to have more input into the hardware underlying their private cloud

Disadvantages

• The cost of a private cloud is prohibitively high for many businesses because of the dedicated hardware required
• Infinite scaling is more difficult to realise, as the virtualisation of servers is still tied to the finite server resources allocated to a private cloud. More servers and hardware need to be purchased whenever the underlying hardware nears capacity