The problem of ‘server sprawl' - and how to tackle it

Tagged with: data centre, server sprawl, efficient it, virtualisation

View Comments and Questions (2)

What's Your Question � Ask Our Experts..

Virtualisation may be one of the most talked-about developments in data centre technology right now - but it is by no means a brand-new concept.

"In fact, virtualisation's legacy stretches right back to the early days of the mainframe, when partitioning was pioneered in order to enable a single machine to run multiple workloads. And it has long been a capability of many high-end Unix machines, too," points out Mandy Shaw, chief strategic architect at Logicalis.

What is new, she says, is that virtualisation can now be applied to the full range of servers in today's heterogeneous data centre, including commodity x86 (or ‘Wintel') hardware.

That will be vital in enabling businesses to tackle the tricky problem of ‘server sprawl', according to Rupert Green, a virtualisation expert at Logicalis. "Quite simply, falling hardware prices has led companies to purchase a new commodity server every time they need to install a new application," he says.

But while those servers may be relatively inexpensive, they still guzzle energy and cooling, occupy valuable floor space, and act as a drain on IT's time and resources.

Worse still, they rarely operate at anything like full capacity. "Many aren't using even 10% of their potential processing power," says Green.

Faced with mounting pressure to deliver new applications at lower cost, more rapidly, and with a higher level of availability, enterprise data centres are turning to server virtualisation to optimise IT resources, he says.

Virtualisation delivers benefits by creating multiple logical servers on one physical machine. This allows businesses to use available server capacity more efficiently by having one server host multiple operating environments, such as Windows or Linux.

That is achieved primarily through the use of software - most commonly from market leader VMware, according to Green, but increasingly from Microsoft (with the Virtual Server) and from the XenSource open source development.

However virtualisation is achieved, its benefit is clear, says Green: to rationalise the sprawl of servers that a company currently needs, to buy, power and maintain, in order to keep its competitive edge.

Your Comments and Questions

Rupert Green, 11 months ago

Hi ... ROI can work work in a number of ways. The most common method is to show a return over a number of years, ususally 3 or 5. This involves comparing the ongoing cost of replacing hardware and the associated hardware maintainance. As most Intel and AMD will be under-utilised the total amount of hardware required to run a given enviroment can be significanly reduced, with the added benefit that the new enviroment is resillient to (physical) server failure. A more sophisticated ROI calculation would take into account the cost of datacenter space, power and cooling and operational costs.