Anyone who spends much time in technology or business circles has heard talk about information or services residing “in the Cloud”, but what exactly does this mean? Is it just a trendy new way to talk about virtualization? For that matter, what exactly is virtualization anyway? What are the advantages and disadvantages of each? How can they save time, money, or both?
Let’s start with the older of the two terms, virtualization. Cloud computing usually utilizes virtualization, but not all virtualized systems and services are in the cloud. Virtualization refers to various techniques used to run multiple servers (a networking concept referring to the digital entity that client machines communicate with on a network) on a single physical piece of hardware. This can be confusing, as the computers that perform these functions are often themselves referred to as ‘servers’, but this is why it’s referred to as ‘virtualization’ in the first place; you’ve got one physical server running multiple ‘virtual’ servers within the network architecture.
Virtualization is popular because it allows a network to spread its load more evenly over the available hardware, and can make fuller use of an individual physical server’s capabilities. It’s scalable to some degree, although this is still limited by how much load can be placed on the network’s hardware.
Cloud Computing often makes use of virtualization to offer its even-more-scalable services. In Cloud Computing, the client rents a certain amount of server “time” or space, although when virtualization is utilized this does not necessarily refer to renting an actual discrete physical server. Companies which offer cloud computing maintain very large banks of hardware–much larger than most clients can cost-effectively support–and can make use of this asset to offer enormous flexibility to their customers. This economy of scale also means that data and services are stored and backed up with a great deal of redundancy, generally at multiple physical sites, meaning much greater security against data loss.
Cloud Computing is not just a service for large companies needing large-scale services. Its inherent flexibility and scalability mean that it can be used for individual consumers’ requirements. Software platforms like Valve’s Steam service, for example, stores all of its users’ applications and application data (usually game software in this case) which can be deleted locally and then retrieved from the cloud at any time. This kind of service obviates the need to keep local backups or physical software installation mediums like DVDs, and saves a great deal of money on shipping, logistics, and other brick-and-mortar retail costs.
Virtualization is, however, not always the ideal solution for cloud computing, and many clients who anticipate consistently heavy demand on their services or who require very high hardware performance at a given moment may wish to lease a dedicated host server. This provides most of the benefits of having an onsite dedicated server, but with things like redundant power sources and automated backups being taken care of by the provider. This is often the best solution for high-traffic websites, for example.
Brenda Panin is an author of this article and a regular contributor to several tech blogs. In her free time she enjoys writing about topics related to web hosting and cloud computing. Useful information for this text has been kindly provided by Ninefold.
Photo courtesy of: cocoate.com