Understanding virtualisation

23rd September 2011

Virtualisation may be a variation on a computing model that’s been around for decades, but this doesn’t make it any easier to understand. As FSN writer Lesley Meall discovers, when she explores some of it’s many incarnations.

There was a time in the dim and distant past when the word computer would have been used to describe somebody who did mathematical calculations (as in the verb to compute). But language is a living thing, and information technology is one of those areas where it seems to be at its most alive. The human computer was eventually replaced by a machine the size of a room, and this computer has evolved into something (a great deal smaller, but) so all-encompassing that its meaning can be determined only with the help of a classifier, such as desktop or notebook – though the meaning of terms such as ‘desktop computer’ and ‘notebook computer’ still tend to depend on who is using them. 

So it is with virtualisation, or should that be virtualization? Most of the time, you can’t figure out what it means without the help of a classifier or context, or both – and even then, the meaning can seem opaque (so the spelling is a minor issue). In its most general sense, the virtualisation of computing is about the creation, provision and use of a virtual (rather than actual) version of something. From a technology point of view, it’s about abstracting computer resources, so that their physical characteristics are hidden from the systems, applications, and users that interact with them. For the non-techie, got more to do with making the most of any computing resources you have (or could have) at your disposal. 

Before you can do this, the widespread use and misuse of ‘virtualisation’ will force you to meaner through a minefield of misnomers, distinguish between theory and practice, and separate reality from hype (on which more, later). But it’s probably worth persevering (to varying degrees), because if virtualisation can live up to its promise, it will repay the effort. ‘Virtualisation helps organisations to cut costs, better utilise assets and reduce implementation and management time and complexity,’ reports Alan Dayley, managing vice president of the infrastructure software team at Gartner, by improving accessibility, performance, reliability, and scalability. 

As all of this would be good news at the best of times, challenging economic conditions can make it seem particularly appealing. However Gartner warns end-user organisations against assessing it in isolation. Dayley suggests the careful building of ‘cost benefit financial models’ that will enable you to ‘fully understand’ the impact of virtualisation, and to compare it with other approaches to IT utilisation, before making any rash leaps of faith. Though as virtualisation can mean so many different things to so many different people in so many different scenarios, you’ll probably need to start out by getting a broad overview of it. 

Unfortunately, just deciding where to start the hunt for enlightenment is a challenge. IDC, Forrester, Gartner and other analysts provide definitions that are variations on a theme, and specialist providers do more of the same (whilst slanting their definitions to emphasise their own products or services). Dan Kusnetsky, who is widely regarded as a virtualisation guru, and is chief analyst and founder of Kusnetzky Group LLC, suggests that virtualisation is easiest to understand if it is broken down into groups: access, applications, processing, storage, and networks – each of which require management and security (which can be virtualised too).


The structure imposed by this diagram can be very helpful, as can Kusnetzky’s definitions, and you can learn more about them with the help of his book Virtualization: A Manager’s Guide. As the title implies, it’s not aimed at those who are heavily into the technology of virtualisation, but it’s not aimed at the total novice either, and whilst it wont turn you into a virtualisation expert overnight, reading it from cover to cover (or just dipping into it) could save you a lot of time (and money) in the long run. It seems to be vendor agnostic and it’s more accessible than a lot of analyst research out there.

Kusnetzky’s guide explains the different types of virtualisation, and then devotes a chapter to each, stating what it is, what it does, and who the main players/providers are in each category, as well as covering important issues such as how layers of virtualisation work and offering some contemporary use cases. The book also provides managers with guidance on when each of these various types of virtualisation is the right choice to make, which could be vital information, as Kusnetzky also points out that ‘using the wrong tool or using the right tool improperly can result in poor performance, higher costs for the organisation,’ and stop it from meeting its objectives. 

He also clarifies the meaning of some of the terms of reference that are often misused in the context of virtualisation; though back in the real world, confusion will remain a distinct possibility. Witness the first Kusnetzky category, access virtualisation, which is also known as desktop virtualisation, which is also known as client virtualisation (as these terms of reference often describe the client-server model, which separates the desktop environment a device can access from the device the resources are on). So this could mean using a smartphone to access resources that wouldn’t otherwise run on it, or using a personal laptop to access resources that are running behind the firewall on a corporate server. 

You may even encounter scenarions where a ‘virtualised desktop’ is used to describe the process of  running mutiple ‘virtual servers’ on a local/remote machine; though the Kusnetzky model describes the creation and/or use of these virtualised servers (which might once have been described as ‘logically partitioned drives’) as processing virtualisation. ‘This covers a broad view, ranging from software than links many machines and makes them appear as a single computing resource to software that allows the resources of a single machine to be presented as several individual machines,’ he explains Kusnetzky. 

Unfortunately, there are contexts where something rather different is meant by the (slightly different) term ‘process virtualisation’, which describes scenarios where technology is used to virtualise processes that would traditionally have been undertaken face-to-face involving real people – and examples range from formal education, through distance learning, to the development of friendships through social networking. Add a classifier, and you get ‘business process virtualisation’, which is more of an umbrella term that’s used to describe the way in which technology can be applied to improve the efficiency of an individual, team, department, or company. 

Which brings us back to where we began: with a word so all-encompassing that its meaning can be determined only with the help of a classifier, or context, or both. So whether you are reading Kusnetzky’s book, trying to decode marketing material, or assessing the impact and implications for finance systems (something FSN will do in a future article), remember that virtualisation is not a single solution or even a set of solutions, and there is a range of interpretations and implementations. There always has been – even eons ago, when IBM originally coined the phrase ‘virtualisation’ (or should that be virtualization) to describe the process we now call outsourcing.