Criticality of the edge puts focus on downtime prevention

Today’s data centers – including those at the edge – have reached utility-like status as a critical public resource, but recent research indicates that criticality isn’t always reflected in robust design and investment to ensure availability at those sites. A study from the Ponemon Institute, commissioned by Vertiv, found just 54% of edge locations are using best practices in design and redundancy. Cost is a factor, with 67% of edge locations saying their risk of an unplanned outage increases as a result of cost constraints. Compounding that fact, 62% of edge facilities say they lack ample resources to bring their network back online if they have an unplanned outage.

Underscoring these findings, 75% of the participants in the Uptime Institute 2020 Global Data Center Survey said their most recent downtime event was preventable. Simply put, despite the increased criticality of today’s data centers, we are seeing more frequent and longer-lasting outages that are preventable with the adoption of availability best practices.

That paints an ugly picture, and the data on outages backs that up. The organizations surveyed in the Ponemon Institute study – all located in the Americas – experienced an average of 31 core data center outages and five edge outages, ranging from individual servers to total system outages, over the past 24 months. Ponemon has been studying data center downtime since 2010, and the average of 31 outages in the 2020 survey was the highest number in the history of the study. The 2020 study was the first to look at edge outages.

The disparity between the frequency of outages in core data centers compared to the edge reflects the size of the facilities and the nature of the outages. A core data center can have thousands of servers, and outages of even a single server are reflected in the total number. Edge facilities are much smaller, with fewer servers, so there is less of a chance for outages.

Preventing data center and edge outages

Looking at the causes of outages in core data centers over the years, most have remained more or less static, but there is one notable outlier. Cybercrime, which barely registered in the first Ponemon Institute report on downtime, was a leading culprit in the 2020 study and the No. 1 cause of outages at the edge. The dramatic increase in outages caused by cyberattacks should inform any strategies for preventing unwanted downtime.

Organizations are deploying increasingly complex, distributed networks that meet the growing demand for computing at the edge of the network, but these sprawling networks also have more access points that present security challenges. This can be addressed with a variety of hardware and software tools that effectively segment the network while still enabling smooth, uninterrupted data transmission. These tools include secure gateways, advanced console servers, secure monitoring, and management software, and even secure, intelligent power and cooling systems that enable safe, remote visibility and control.

There are a variety of UPS systems and architectures, and organizations should assess the load and the criticality of their different IT systems, define the level of redundancy needed, and choose their UPS systems accordingly. Likewise, there are more options for UPS batteries now than ever before, with lithium-ion emerging as an increasingly viable alternative to traditional valve-regulated lead acid (VRLA) batteries. Lithium-ion batteries last longer, deliver longer runtimes, function better at higher temperatures, and as adoption has increased in recent years, the return on investment has shrunk to a single refresh cycle. While early lithium-ion options were limited to larger installations in core data centers, today there are a variety of UPS systems for edge environments that support lithium-ion batteries.

There was a time when edge deployments were considered something less than mission critical. As more and more applications move to the edge, however, that has changed. Connectivity and availability across the network are indistinguishable. Today’s edge is more sophisticated and critical than the IT closets of the past and requires an equally sophisticated infrastructure befitting that criticality.

About the author

Martin Olsen is the vice president of global Edge and integrated solutions at Vertiv. 

DISCLAIMER: Guest posts are submitted content. The views expressed in this blog are that of the author, and don’t necessarily reflect the views of Edge Industry Review (EdgeIR.com).

Article Topics

data center management  |  edge data center  |  outage  |  power management  |  Vertiv