Cloud hosting services provide hosting for websites on virtual servers which pull their computing resource from extensive underlying networks of physical web servers. It follows the utility model of computing in that it is available as a service rather than a product and is therefore comparable with traditional utilities such as electricity and gas. Broadly speaking the client can tap into their service as much as they need, depending on the demands of their website, and they will only pay for what they use.
Rather than being hosted on one single instance of a physical server the website is hosted on a virtual partition which draws its resources, such as disk space, from an extensive network of underlying physical servers. If one server goes offline, it dilutes the level of resource available to the cloud a little but will have no effect on the availability of the website whose virtual server will continue to pull resource from the remaining network of servers. Some cloud platforms could even survive an entire data centre going offline as the pooled cloud resource is drawn from multiple data centres in different locations to spread the risk.
The underlying physical servers are still housed within data centers and so benefit from the security measures that those facilities implement to prevent people accessing or disrupting them on-site.
Resource is available in real time on demand and not limited to the physical constraints/capacity of one server. If a client’s site demands extra resource from its hosting platform due to a spike in visitor traffic or the implementation of new functionality, the resource is accessed seamlessly. Even when using a private cloud model the service can often be allowed to ‘burst’ to access resources from the public cloud for non-sensitive processing if there are surges in activity on the site.
The client only pays for what they actually use. The resource is available for spikes in demand but there is no wasted capacity remaining unused when demand is lower.
Load balancing is software based and therefore can be instantly scalable to respond to changing demands