When asked to picture the typical data centre, many will conjure an image of a windowless room filled with hi-tech computers, a multitude of wires and the audible hum of servers processing information quicker than it takes to blink; a box room that has the ability to power whole cities.
For many facilities managers and engineers, their main responsibility is to ensure the data centre delivers performance in line with business demands, so their focus tends to be on understanding how any change made inside the data centre environment will affect overall performance. Whilst managing risk and being aware of the influence of each internal dynamic of the data centre is undeniably important, one critical influencing factor traditionally gets overlooked - what’s going on outside the box?
‘Thinking outside of the box’ is a saying that has been around for years and when it comes to data centre management, it’s a phrase that needs to be used more often. There are a multitude of external factors which can impact the successful running of a data centre and are arguably as important to consider as the internal white space dynamics themselves. So what external factors need to be considered and how important are they when making decisions that benefit the efficiency and effectiveness of the data centre?
What makes a data centre sweat?
Often, areas that would appear be a great fit for data centres in terms of location and space can turn out to be problematic due to environmental extremes. Singapore, by example, is currently enjoying a huge expansion in data centre operations due to its proximity to China, political stability and its business friendly regulations. Colocation alone was valued in excess of $1 billion in revenues in 2014 with expectations of it reaching around $1.2 billion by the end of this year. However, the climate is hot and humid, minimising the amount of free cooling and adding considerable cost on the cooling infrastructure.
In addition, another important factor that can affect data centre performance is its proximity to surrounding infrastructure. With the industry focusing on utilising more fresh air for cooling, the air quality as well as its temperature and humidity is key to the data centre performance. Contamination from neighbouring industry, smoke from forest fires or desert sandstorms all pose risks to air quality and hence the cooling performance and white space environment.
Furthermore, the cooling plant requires significant volumes of fresh air to ensure correct performance. Restrictions to air supply either from neighbouring buildings or due to noise proofing can starve chillers or cooling towers of fresh air, forcing the units to recirculate hot exhaust air and suffer significant performance deterioration.
These subtle but fundamental connections between inside and outside of the box are the key challenges facing data centre designers and operators. However, these external effects can be managed. Data centre managers have long made use of engineering simulation - specifically, 3D modelling, power system simulation (PSS), and computational fluid dynamics (CFD) to predict cooling within the four walls of the data centre. Simulation provides a safe, offline environment to test any potential change, without fear of failure or downtime. It also allows Facilities Managers to optimise data centre performance in terms of efficiency, resilience and conformance. However, whilst this kind of 3D modelling is well established for the internal data centre environment, it can also be used to predict and tackle any external influences on the data centre.
For instance, when looking at the Singapore environment, engineering simulation could be used to assess the impact of high temperatures and humidities on the data center performance. By simulating the environment, data centre designers and owners can push the number of free cooling hours achievable without risking critical IT applications.
When it comes to assessing the impact of external infrastructures built near existing DCs, the use of 3D modelling would allow engineers to predict exactly the impact on performance even as early as the site selection stage. This critical view of the future offers the opportunity for proactive steps to mitigate any potential problems The alternative of reacting to issues further down the line are ultimately more restrictive, take longer and are significantly costlier for the business.
The use of predictive simulation to understand and predict the effects that external influences will have on a data centre could allow facility managers to truly understand and prevent all risks to the effective running of the DC. By taking care of outside of the box, the business can ensure that their critical applications will continue to operate inside the box.
Jon Leppard, Director at Future Facilities
Image source: Shutterstock/Scanrail1