Frequency and Severity of Data Center Outages Increases, Says Uptime Institute

Data center operators have, by and large, succeeded in lowering energy waste – as the average Power Usage Effectiveness (PUE) was reported at a record average of approximately 1.58 (although year/year improvements are decreasing), according to the findings of Uptime Institute’s eighth annual Data Center Survey. In alarming contrast, the survey results also revealed that both the rate of outage occurrences and the severity of each outage have increased from those reported just a year ago.

More than 80% of respondents said their outage(s) were preventable – leading causes of downtimes consisted of human error, power outages, network failure, configuration errors and third-party provider outages.

While nearly two-thirds of respondents (61%) said having workloads spread across on-site, colocation, and cloud deployments have made their overall IT more resilient, this was not supported by the evidence, stated Uptime Institute. The number of respondents that experienced an IT downtime incident or severe service degradation in the past year (31%) increased over the 2017’s survey (by about 25%). And in the past three years, almost half (48%) of our 2018 survey respondents had an outage in their own site or service provider. According to a research report by the Ponemon Institute, a data center outage costs around $9,000 per minute.

This global Uptime Institute data center industry survey is conducted annually. This year, the survey was conducted between February and May 2018 and includes responses from nearly 900 data center operators and IT practitioners affiliated with enterprise and service provider facilities in more than 50 countries worldwide.

Many survey respondents struggle to assess the business case and effectiveness of their hybrid IT architectures (defined as any mix of on-premises data center capacity and off-premises resources such as colocation, cloud, hosting and XaaS). Only about half are confident they understand the true costs and risk/performance tradeoffs of their chosen approach.

Edge Computing Capacity

The Uptime Institute survey data also showed that crucial to the growth of the distributed digital infrastructure is the implementation of highly distributed and/or edge computing, enabling more efficient operations through automation, real-time data analysis and artificial intelligence (and machine-learning) control associated with remotely-managed facilities. More than 40% of respondents stated they expect their organization will require edge computing capabilities, which was defined as “requirements that will necessitate processing data closer to the source of its generation/use”.

“Today, operators are grappling with new challenges, including increased complexity and high interdependency of systems and data centers,” said Andy Lawrence, Executive Director of Research at Uptime Institute. “Looking ahead, many are expecting to deploy significant new hybrid and edge computing capacity, which will support new services but will add an additional layer of complexity in doing so. Edge computing is exciting because of the improved performance and scale it can offer to nnext-generationtechnologies like artificial intelligence, the Internet of Things and even autonomous driving applications. We expect to see substantial growth in the edge over the next few years. Edge has the ability to keep building upon each set of application improvements and the advances of previous versions, which will cause rapid improvement in capability development and implementation.”