Data center staff are getting burnt out – and this could be causing more frequent outages

Data center worker in a server room working on racks.
(Image credit: Getty Images)

Human error is one of the biggest causes of data center outages, and long shifts for staff could be at least partly to blame.

According to research by Uptime Institute, about half of data center operators have suffered an outage in the past three years. Across its 25 years of data, Uptime estimated that human error - directly or indirectly - contributed to somewhere between two-thirds to four-fifths of all incidents.

Almost half of those outages were caused by staff failing to follow procedures.

However, Uptime has said that one under-appreciated factor could be fatigue – the result of the long shifts that data center staff can often find themselves scheduled onto. Fixing the problem might be a challenge, though.

“The relationship between shift length, fatigue and human error is well documented, but less clear is how the data center industry can define shifts that help minimize human error,” said Rose Weinschenk, a research associate at Uptime.

That’s because, Weinschenk said, best practice from other industries does not always translate well into the data center world, where 24-7 service availability is the expected standard. Instead, data center bosses who want to change shifts to limit fatigue may find themselves having to navigate the demands of employees and particular issues in different regions.

Staff in data centers certainly have plenty to keep them busy. They might be installing new racks of hardware or carrying out other requests involving the computing hardware stored within, like upgrading servers, or updating software or networks.

Data center workers are also tasked with inspecting and testing infrastructure to make sure that cooling or power or security systems are operating as they should.

That may involve troubleshooting, repairing or replacing equipment to keep the data center operating at a high level, and generally responding when things go wrong.

A data center technician can expect to earn from around £40,000 a year upwards depending on experience and location. Staffing can often be one of the biggest ongoing costs of a data center.

Fatigue is a growing problem for data center staff

Uptime pointed to broader academic research which shows staff working more than 10-hour shifts are significantly more likely to experience fatigue. The risk of workplace injury due to fatigue-related accidents across a range of industries is 15% higher in 10-hour shifts than eight-hour shifts - and jumps to 38% higher for 12-hour shifts.

While eight-to-10 hour shifts are most common in the data center industry, Uptime’s own research found there are some geographic variations. Around 17% of all respondents to its survey reported shifts of more than 10 hours, and in Asia-Pacific this stood at 22%. 

Of respondents from Europe, only 13% said they operated shifts of more than 10 hours, but they reported three times as many five- to seven-hour shifts as respondents from Asia-Pacific.

Part of this is down to regulatory climates, Uptime said. In some European countries, labor laws do not allow night shifts to exceed eight or 10 hours as standard, which may lead companies operating there to hire more part-time employees to make up their staffing shortfall.

RELATED WHITEPAPER

Doing that in the US, however, is harder because there is no statutory obligation for the employer to provide healthcare coverage if employees work less than a 40-hour week, which means workers are reluctant to reduce their hours.

Cutting long night shifts – which have the greatest risk of workplace injury – might make sense, but doing so is not without its own problems.

Having staff that only work weekends or nights – rather than rotating through shifts – can make for monotonous work and hurts team cohesion.

Some staff also prefer 12-hour shifts over several days, because it can bring additional overtime pay and longer blocks of time off work. Getting staff to cover gaps in rotas with relief shifts also risks them coming back to work still tired, putting them at risk of making the same mistakes you were trying to avoid.

Adjusting shifts for existing employees could hurt morale and – at least in the short term – lead to higher levels of fatigue as staff adjust to their new schedule.

How to support data center staff

Despite these challenges, Uptime said that data center operators should avoid shift lengths of more than 12 hours and should make sure staffing levels and schedules minimize abnormally long shifts.

Managers should monitor overtime and rest periods between shifts to avoid calling in already-tired staff. And while bosses should take into account employee preferences, they also need to remember that shift workers can ignore potential risks to their own job performance and health when requesting a particular work pattern.

More broadly, data center teams are struggling to attract and retain qualified staff. Uptime has previously reported that data centers still employ shockingly few women.

The average proportion of female employees at respondents’ organizations is just 8% — lower, it noted, than in industries such as construction, mining and manufacturing.

Steve Ranger

Steve Ranger is an award-winning reporter and editor who writes about technology and business. Previously he was the editorial director at ZDNET and the editor of silicon.com.