13 June 2009

Mark Wilson and Dr Bernd Kosch of Fujitsu Technology Solutions discuss the strategy behind virtualisation.

Once a niche technology for test and development environments, virtualisation has moved into the mainstream as organisations embrace the benefits in efficiency and flexibility that it offers. But virtualisation in itself is neither a quick fix for complex IT environments nor a guaranteed source of bottom-line benefits. It all depends on how you manage the technology and how you adapt the processes and culture of your organisation to new ways of working.

Virtualisation scores high on every CIO’s must-do list. The potential advantages to be gained from breaking the fixed connections between software and hardware is galvanising IT managers and architects across all kinds of large organisations. A virtualised IT infrastructure is capable of providing increased efficiency, enhanced flexibility, and improved quality of service. Now that the technology has come of age and the concept is on decision makers’ agendas, IT teams are getting to grips with the reality of implementation. However, sticking to a set of key guidelines will help to recognise and head off the challenges.

Virtualisation promises a much closer match between the demands of applications and the resources that serve them

Mark Wilson and Dr Bernd Kosch, Fujitsu Technology Solutions Touching down in a virtual world
Consolidating multiple servers into a smaller set certainly results in significant infrastructure cost savings. Power consumption and cooling efforts will reduce dramatically – particularly if the underlying hardware is replaced with powerful yet energy-efficient new models. One Fujitsu customer in the finance sector is saving around two million kW/h a year (€200k) by moving 1,000 servers to a virtualised environment using Primergy servers. For capacity planning it has to be taken into account that there are still some applications that continue to demand the use of dedicated servers.

Unleashing new demand
Virtualisation promises a much closer match between the demands of applications and the resources that serve them. Customers and managers alike gain greater visibility of the relationship between delivery to a business process and the costs of that delivery. In addition, fulfilment of customer requirements becomes much faster in a virtualised world. FlexFrame for Microsoft environments – an infrastructure solution based around Microsoft System Center products with an integrated hardware management element – is automatically providing incremental resources for running applications in minutes rather than hours or days. The flip side of this greater speed of delivery can be unexpected: a surge in demand as user perception switches from viewing IT as a scarce resource that’s often difficult to implement to regarding it as a free and limitless capability.

The scenario we’ve looked at up to this point is restricted to the impact of virtualisation on data centre operations. But what if the technology is applied to the desktop as well as the back-end infrastructure? Extend virtualisation to the desktop using virtual desktop infrastructure technologies or software streaming and the demand for power at the data centre will grow even faster and higher. Network usage will also grow in orders of magnitude. There are management issues too. If you have virtual desktops running on servers in the data centre, who manages them – the desktop support team or the server support team? Or do these teams merge? Also, who manages the network when it is no longer entirely made up of physical cables and switches but encompasses a virtual network infrastructure too?

By consolidating physical assets into the data centre and removing servers from office environments, IT managers can create a dangerous illusion of control. The equipment may all be in one place but that doesn’t mean it is being centrally managed. Management doesn’t emerge by itself: it has to be applied. The organisation’s virtual machines are still islands of capability, each of which requires active monitoring and management. And if those capabilities weren’t managed in the physical world, virtualising them will not supply the remedy.

Think of virtualisation as a strategy rather than a set of technologies. Adapt your existing management processes and develop new ones where necessary, both to enable and contain the creative potential of virtualisation. There are a number of key touchstones for ensuring success with your virtualisation strategy. One: make a full and faithful inventory of what your servers do and how they relate to each other. Two: build a provisioning model with supporting workflow to allow the cost of a new virtual machine to be attributed to an individual or department. This cost can be real or indicative – but it should be meaningful. Three: design a lifecycle mechanism to ensure that storage is not filled with duplicate, unused virtual machine files. Four: continue to perform capacity planning and load placement in order to find workloads that complement each other. And five: record appropriate metadata for each virtual machine, including details of its owner, dependencies, and SLA level.

Since the only certainty in business is that tomorrow will look very different from today, virtualisation’s potential to reconfigure IT delivery to match needs and resources is a welcome contribution to the management task. But it’s just that: a contribution, not a replacement.

Mark Wilson is the technology lead for Microsoft at Fujitsu Services.

Dr Bernd Kosch is Green IT advisor at Fujitsu Technology Solutions.

This article first appeared in the Summer 2009 issue of Finance on Windows.

Add a comment