The strategic importance of the data centre can be in no doubt. Businesses are evermore reliant on the IT infrastructure that automates and supports their mission-critical processes, from order placing to payroll. But today, the data centre is undergoing a radical metamorphosis, and the ability to understand and control this change places a heavy burden on today’s IT management teams.

The need to impose order on the IT infrastructure is reflected in the push to centralise or, at the very least, to rationalise operations. Infrastructure is being moved from branch and regional offices to central sites, and those data centres themselves are being consolidated into a few units.

Alongside efforts to centralise the IT infrastructure, the variety of applications being hosted in data centres is also growing and becoming more complex to manage. One example has been the drive towards voice-over-Internet Protocol communications: in many ways, digital voice traffic becomes ‘just another application’ sitting on the network, but it is an application with some very specific requirements, not least of all the need to minimise latency.

To manage such mixed environments, businesses have historically relied on ‘point solutions’, notes IT advisory group Forrester Research. But recent advances in remote management tools have enabled businesses to “intelligently collapse similar services”, to reduce the management burden and still offer full-service sites, says Robert Whitely, a senior analyst at Forrester.

As this happens, business leaders can get a clearer insight into the operational costs of these consolidated units far more easily than tracking expenditure across infrastructure dispersed over numerous locations. However, such options do not get away from a harsh reality: operating costs are rising fast.

The cost of running a data centre has traditionally been associated purely with the capital outlay for the building and hardware. However, as ‘high density systems’ – such as blade servers – have been introduced, the cost of the physical space required to house the servers and other equipment has taken second place to the cost of powering the servers and their associated cooling systems. According to analyst group Gartner, today’s energy costs form less than 10% of the overall IT budget; within a few years, that could rise to 50%. And in the UK, data centre electricity bills have been rising at 30% a year for three straight years.

The volatility of energy prices is not the only issue; the reliability of supply is also a concern. Domestic power production in the UK is set to drop in coming years, with roughly one third of the UK’s current generating plants due for retirement by 2025. There is, as yet, no large scale building programme underway to address that shortfall.

For many, the growing power crisis is likely to intensify efforts to make better use of the data centre infrastructure. To date, efforts have largely focused on server virtualisation, which can boost utilisation rates. But the complexity of today’s consolidated data centres requires a wider approach than just managing server utilisation. The applications running in a data centre can have hugely different requirements: email needs relatively low levels of processing power, but eats into storage pools; the opposite is true of enterprise resource planning systems; and the arrival of PC blades or VoIP systems into the data centre adds greater complexity to workload management.

According to analyst group IDC, the pressure these changes are bringing to bear on data centre operations is hastening the drive to an IT service management model, where policies are “implemented by intelligent rules-based management tools”.

Further reading

• View from the centre
  Data centre consolidation is reshaping the challenges of managing the modern IT infrastructure, Information Age’s latest survey shows.
  

• Gravitational pull
  Businesses are looking to impose control over their dispersed IT assets.