In some respects cloud servers work in the same way as physical servers but the functions they provide can be very different. When opting for cloud hosting, clients are renting virtual server space rather than renting or purchasing physical servers. They are often paid for by the hour depending on the capacity required at any particular time.
Traditionally there are two main options for hosting: shared hosting and dedicated hosting. Shared hosting is the cheaper option whereby servers are shared between the hosting provider’s clients. One client’s website will be hosted on the same server as websites belonging to other clients. This has several disadvantages including the fact that the setup is inflexible and cannot cope with a large amount of traffic. Dedicated hosting is a much more advanced form of hosting, whereby clients purchase whole physical servers. This means that the entire server is dedicated to them with no other clients sharing it. In some instances the client may utilise multiple servers which are all dedicated to their use. Dedicated servers allow for full control over hosting. The downside is that the required capacity needs to be predicted, with enough resource and processing power to cope with expected traffic levels. If this is underestimated then it can lead to a lack of necessary resource during busy periods, while overestimating it will mean paying for unnecessary capacity.
With cloud hosting clients get the best of both worlds. Resource can be scaled up or scaled down accordingly, making it more flexible and, therefore, more cost-effective. When there is more demand placed on the servers, capacity can be automatically increased to match that demand without this needing to be paid for on a permanent basis. This is akin to a heating bill; you access what you need, when you need it, and then only pay for what you’ve used afterwards.
Unlike dedicated servers, cloud servers can be run on a hypervisor. The role of a hypervisor is to control the capacity of operating systems so it is allocated where needed. With cloud hosting there are multiple cloud servers which are available to each particular client. This allows computing resource to be dedicated to a particular client if and when it is necessary. Where there is a spike in traffic, additional capacity will be temporarily accessed by a website, for example, until it is no longer required. Cloud servers also offer more redundancy. If one server fails, others will take its place.
Below are the key benefits of cloud servers:
Flexibility and scalability; extra resource can be accessed as and when required
Cost-effectiveness; whilst being available when needed, clients only pay for what they are using at a particular time
Ease of set up; Cloud servers do not require much initial setup
Reliability; due to the number of available servers, if there are problems with some, the resource will be shifted so that clients are unaffected.
For more information on the different version of cloud servers please visit the following pages:
What is a Public Cloud?
What is a Private Cloud?
What is a Hybrid Cloud?
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.
It exists as an alternative to hosting websites on single servers (either dedicated or shared servers) and can be considered as an extension of the concept of clustered hosting where websites are hosted on multiple servers. With cloud hosting however, the network of servers that are used is vast and often pulled from different data centres in different locations.
Practical examples of cloud hosting can fall under both the Infrastructure as a Service (IaaS) and Platform as a Service (PaaS) classifications. Under IaaS offerings the client is simply provided with the virtualised hardware resource on which they can install their own choice of software environment before building their web application. On a PaaS service however, the client is also provided with this software environment, for example, as a solution stack (operating system, database support, web server software, and programming support), on which they can go straight to installing and developing their web application. Businesses with complex IT infrastructures and experienced IT professionals may wish to opt for the more customisable IaaS model but others may prefer the ease of a PaaS option.
A development of the concept of cloud hosting for enterprise customers is the Virtual Data Centre (VDC). This employs a virtualised network of servers in the cloud which can be used to host all of a business’s IT operations including its websites.
The more obvious examples of cloud hosting involve the use of public cloud models - that is hosting websites on virtual servers which pull resource from the same pool as other publicly available virtual servers and use the same public networks to transmit the data; data which is physically stored on the underlying shared servers which form the cloud resource. These public clouds will include some security measures to ensure that data is kept private and would suffice for most website installations. However, where security and privacy is more of a concern, businesses can turn towards cloud hosting in private clouds as an alternative - that is clouds which use ring-fenced resources (servers, networks etc), whether located on site or with the cloud provider.
A typical cloud hosting offering can deliver the following features and benefits:
Reliability; 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
Physical Security; 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
Scalability and Flexibility; 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.
Utility style costing; 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.
Responsive load balancing; load balancing is software based and therefore can be instantly scalable to respond to changing demands