With many storage and hardware failures caused by environmental factors, it is no surprise that data centres are now routinely monitored for temperature and humidity. But what about other environmental factors like dust? With most data centres enduring a constant flow of people, equipment and packaging, carbon and concrete dust also cause problems – as do rodents and leaks from air conditioning equipment.
With investments in data centres continuing to increase, can a data centre manager risk not only downtime but also invalidating data centre equipment warranties and undermining business confidence? In this article I’d like to explain the importance of extending continuous monitoring to embrace the cleanliness of the data centre and outlines issues that need to be considered to minimise the impact of environmental factors, from regular cleaning to proactive changes to working practices.
Over the past few years, attitudes to data centres have changed radically. Driven in part by the growth of co-location centres but also the move to reduce power and improve environmental performance, organisations now have high expectations of the reliability and performance of data centre technologies.
As such, investment in data centre design is now paramount, leveraging innovation in cooling technologies to reduce costs and drive down power consumption. Companies are also investing in continuous monitoring solutions to ensure optimum temperature and humidity at all time, to maximise performance and minimise the risk of failure.
Yet there are other environmental factors that also affect data centre performance, most notably cleanliness. With a stream of in house technicians and external consultants undertaking new cabling, server and rack upgrades, these so called clean data centre environments are under constant pressure from dust and dirt to forgotten tools, screws, cage nuts, cable ties and packaging.
Today it is estimated that just 50% of data centres are ever cleaned – and that is often a one-off event following a major refurbishment such as a client visit or a senior management visit. So what are the risks associated with this practice and how can an organisation create a better environment without embarking upon a major investment?
5 reasons your data centre should be cleaner
- Improve performance and availability: With the growing dependence upon the cloud and 24×7 operations, data centres are mission critical operations and data centre managers are required to explore any opportunity to reduce failure and minimise downtime. By reducing the hazards associated with dust and debris, regular clinical cleaning in association with humidity and temperature monitoring is proven to reduce the failure rate and increase reliability.
- Business confidence: In far too many organisations a brand new data centre can move from pristine state to de facto storage room in a matter of months. Overrun with packaging left behind from a new equipment install – with or without equipment – is not only a fire hazard; it also undermines management confidence in the quality of all data centre processes.
- Valid warranty: Dust and dirt may not be overly visible but should equipment failure occur the warranty may well be invalid if the equipment is discovered to be full of dust.
- Competitive position: For co-location sites, continuous monitoring technologies are a key selling point. Extending that model to include clinical cleaning not only minimises the risks associated with equipment failure – and hence invoked Service Level Agreement (SLA) penalties – but also provides confidence to clients and prospects in the rigour of the operation.
- Safeguard investment: During 2011-2012, the UK invested an estimated $3.35 billion in data centres, the second highest spending of any country, according to the Datacentre Dynamics Global Industry Census 2011. Ensuring the monitoring techniques and working practices are in place to safeguard these environments is key to maximise the value of this investment.
So what is the way forward for organisations looking to minimise the risk associated with data centre performance?
5 steps to a cleaner data centre
- Needs analysis: The first step is an initial audit to establish the extent of the problem and the required solution. Typically an organisation will have a deep clinical clean to remove the dust particles, static and any other identified contaminants from the room, with all areas treated with anti-static solutions. This can then be followed up be regular cleans on a quarterly or biannual basis.
- Professional expertise: Clinical cleaning is a skilled operation requiring understanding of data centre equipment and structure, including sub-floor and ceiling ducts, as well as the specialist cleaning equipment. It is essential the cleaning is undertaken in a way to ensure continuity and minimise any impact on day to day activity – typically at night or over a weekend.
- Working practices: Obviously there is no need for data centre staff to adopt biohazard suits. But basic good practice – such as removing all packaging and tools after use – and ensuring the wheels of any trolleys use to carry in equipment are clean, are important. By ensuring the data centre space is reserved only for core, business critical systems, an organisation can foster an attitude that reflects its importance.
- Pragmatic approach: Once a deep clean has been undertaken – often after a major recabling or refit exercise – simple maintenance will suffice. A basic quarterly or biannual clean will minimise risk and ensure warranties are valid without requiring a massive investment.
- Leverage expertise: Data centres typically include many areas that are not visited on a day to day basis, allowing problems to arise unseen. Using the same expert team to deliver the regular clean provides organisations with an additional set of eyes and ears to highlight evolving data centre risks – from signs of rodents to equipment leaks and holes in the walls that may be undermining the investment in innovative cooling designs and technology.