Uptime Institute's 10th Annual Data Center Survey is here!

It’s HERE!!! Every year, Uptime Institute reaches out to thousands of industry leaders, enterprises and suppliers to ask them about their view of where the data center and digital infrastructure industry is going and what kinds of challenges they are dealing with today and expect in the future. We ask about trends and migrations, tactics and strategies, challenges and successes. This survey is the most comprehensive and longest-running of its kind and used by thousands of companies worldwide to influence their own IT choices and directions. The results may also make you think about some of your long-held ideas which may be becoming widely accepoted or perhaps no longer accurate or strategic.

This year, more than 1300 people responded. The 10th annual survey was conducted in the spring of 2020 and the results provide an overview of the practices, experiences and underlying trends in the mission-critical digital infrastructure industry, today and in the future.

I encourage you to listen to the entire narrative by Andy Lawrence, our head of our Uptime Intelligence research group. Andy will describe a sector that is grappling with a number of difficult issues including the increasing scope and frequency of outages, the actual rate at which migration to cloud is occuring, increasing complexity of hybrid infrastructures and their resilience and the resulting performance expectations.

The survey also confirms that it is an industry that is growing in absolute size and one that is adapting to rapid change on multiple levels. In almost every area under discussion — outages, resiliency, staffing, placement of workloads, deployment of innovation and the use of cloud — there is considerable diversity in the strategies being employed by the professionals chartered to deliver business critical results. And perhaps most importantly, respondants overwhelmingly agree that they have the ability to address each of these challenges IF THEY CHOOSE TO take the required steps and actions! (i.e. Outages are preventable if more attention is paid to operational planning, staffing shortages can be addressed if they widen their searches, data centers will be able to withstand more stress if they were designed and verified properly, etc)

Key findings:

The enterprise data center is neither dead nor dying. The migration of critical loads to a public cloud is happening slowly, with more than half of workloads expected to remain in on-premises data centers in 2022.
Transparent clouds are good for business. Cloud operators would win more mission-critical business if they were more open. Enterprises want greater visibility into facilities and how resiliency is achieved.
Edge is still on the edge. Most organizations expect their edge computing requirements to increase somewhat in 2020, but fewer than 20% expect a significant increase.
Average site energy efficiency has flatlined. Power usage effectiveness values have not improved much across the industry since 2013. But because more work is now done in big, efficient facilities, the overall energy efficiency of IT has improved.
Rack densities are rising, but facilities are not stretched. The mean average density for 2020 was 8.4 kilowatts per rack. Densities are rising, but not enough to drive wholesale site-level changes in power distribution or cooling technologies.
Bigger outages are becoming more painful. Outages generally continue to occur with disturbing frequency, and the bigger outages are becoming more damaging and expensive — a fact supported by Uptime Institute survey findings for three years running.
Operators admit most outages were their fault. Three-quarters of respondents admit that, in hindsight, their most recent major outage was preventable. With more attention and investment, outage frequency would almost certainly fall significantly.
Power problems are still the biggest cause of major outages. Systems/software and networks may be catching up, but power failures — which impact everything on-site and can cause knock-on effects — are the most likely cause of major outages.
Hardware refreshes are less frequent. Operators are upgrading or replacing their servers less frequently. However, the slowdown in Moore’s law means the potential energy savings from frequent refreshes are no longer very significant.
The data center staffing crisis is getting worse. The portion of managers saying they have difficulty finding qualified candidates for open jobs has risen steadily over the past several years.
Artificial intelligence won’t take over … yet. Artificial intelligence and automation will not reduce data center operations staffing requirements in the next five years, according to the majority of respondents. After that, however, most think it will.
Water use is unmetered by many. Despite the growing threat of water scarcity, only half of respondents say their organization collects water usage data for their IT/data center operations.
More work is needed to address the workforce gender imbalance. The proportion of women in the data center industry remains very low. Despite pressure and good intent, relatively few operators have a plan or initiative in place to boost the hiring of women.
Use of availability zones is now mainstream. The use of multi-data center availability zones is now common beyond hyperscale operators, with half of respondents saying they use this approach.