Data center downtime is no joke. It can literally make the difference between a data center surviving and failing. And a new study by the Ponemon Institute shows that modern data centers and data centers at the edge are more susceptible to downtime than ever before. This is because data centers are much more complex than they ever have been. Most core data centers suffer 2.4 facility shutdowns per year, and some of those last around 138 minutes – more than two hours! Edge computing data centers experience twice as many shutdowns, but that average half the duration of core data center outages.
In addition, it is helpful to remember that although total facility failures occur with the least frequency, individual server or rack failures can also be costly, especially in Edge data centers, where every piece of equipment has some critical function.
At the outset it is also important that we define core data centers and edge data centers. Edge data centers are usually about ⅓ the size of their counterparts, although the term edge does not refer to size. Edge refers more to the data center location, generally closer to where the data center is needed to increase speed and response times, and save bandwidth.
Jump to Section:
- Conflicting Priorities
- The Cost of Downtime
- How to Prevent Data Center Downtime
- There is Room for Improvement
Data center managers are faced with decisions about efficiency, the transition to Net Zero carbon emissions, and avoiding redundancies whenever possible, but this also can leave them susceptible to downtime events if a problem occurs. This is illustrated by the causes of downtime: UPS battery failures, human error, equipment failures, other UPS equipment failures, and cyberattacks.
Respondents to the survey revealed that over half (54%) are not using best-practices, and that risks of data center downtime are increased because of cost concerns.
The Cost of Downtime
While cost concerns often increase the risk of downtime, the actual cost of downtime can be much greater. According to a 2014 survey by Gartner, facility downtime cost an average of nearly $5,600 per minute, or between $140,000 to over half a million dollars per hour depending on the organization size. These costs continue to rise, with more recent statistics from the Ponemon Institute survey mentioned above calculating average costs at nearly $9,000 per minute.
It’s more than just the money costs though. The real cost comes in reputation and customer service. Data centers that suffer above average downtime are much more likely to go bankrupt. Uptime is perhaps more critical than it ever has been, and customers remember problems far more easily than they remember reliable service over time.
So what do we do to prevent data center downtime?
How to Prevent Data Center Downtime
There are solutions to downtime issues, and many are known to data center managers. However, they are easier said than done. Here are a few of them:
- Adopt best-practices – The fact that most data centers know they are not following best-practices reveals they know what to do, they are just not doing it.
- Invest in new equipment – Equipment failures come in outdated equipment not up to the current needs of the data center. Replacing it is one of the easiest ways to reduce or eliminate downtime.
- Improve your training – Be sure that all employees, both existing and new, are aware of best practices and what you expect of them on the job. Make training comprehensive and focus on outcomes and skills that build long-term success.
- Improve your documentation – Your data center plans, including power, cabling, cable management plans, and others should be thoroughly documented and available to employees. If not, in the words of Captain Picard, “Make it so.”
- Don’t fight redundancy – Redundancy is a good thing for the most part. You certainly don’t want to overdo it, but you do need to have contingency plans and equipment in case downtime does happen.
Of course, these solutions are simplified, nor are they always possible for data center managers to achieve with the resources they have available.
There is Room for Improvement
The takeaway from this data is twofold. First, data center downtime at these rates are unacceptable for most organizations. The second is that there are solutions, and there is plenty of room for improvement. Among the solutions mentioned above, there are some critical elements.
- Redundancy – This has been preached from the beginning for both core and edge data centers, yet half of data centers have issues in this area. As a result, there is a trend toward more redundant equipment, especially at the edge, as large and small operations seek to better manage data center downtime.
- Remote monitoring systems and AI – The other advancement that seeks to solve the issue of human error and detect equipment issues before they become a problem is remote monitoring and AI. Machine learning can help data center managers fix issues before downtime occurs, and helps them respond faster when a problem does occur.
Simply adding these two things can take data centers a long way toward greater uptime and more reliable service. After all, this is the goal of both core and edge computing.
Whether you manage an existing data center or you are considering starting one from scratch, we here at AnD Cable Products are here for you. We can help you with everything from cable and rack management to labeling systems and remote monitoring. Have questions? Contact us today. We’d love to start a conversation about your specific needs.
Physical Layer Environment Network Security Monitoring and Control
Full visibility, network security and control of your physical layer environment. Monitor your entire hybrid cloud and IT infrastructure from a cloud-based, integrated dashboard:
- Introducing the A150 System
- A150 System Architecture – High-Level Overview
- A150 System Features
- System Controller Hardware and Specifications
- Monitoring Controllers, Probes and Sensors
About the Author
Louis Chompff, Founder & Managing Director, AnD Cable Products
Louis established AnD Cable Products – Intelligently Designed Cable Management in 1989. Prior to this he enjoyed a 20+ year career with a leading global telecommunications company in a variety of senior data management positions. Louis is an enthusiastic inventor who designed, patented and brought to market his innovative Zero U cable management racks and Unitag cabel labels, both of which have become industry-leading network cable management products. AnD Cable Products only offer products that are intelligently designed, increase efficiency, are durable and reliable, re-usable, easy to use or reduce equipment costs. He is the principal author of the Cable Management Blog, where you can find network cable management ideas, server rack cabling techniques and space saving tips, data center trends, latest innovations and more.
Visit https://andcable.com or shop online https://andcable.com/shop/