Mark Gaydos is Chief Marketing Officer for Nlyte Software.
According to Fortune Magazine, “Cyber Monday 2016 was the biggest day in the history of U.S. e-commerce. Consumers spent $3.45 billion online…”
Also, Practical Ecommerce reports that “consumers spent a total of $12.8 billion online in the U.S. during the five-day period from Thanksgiving Day through Cyber Monday, 2016”. Black Friday 2016 was the first day to generate more than one billion dollars in online sales from mobile devices.
Black Friday, the day after Thanksgiving Day in the U.S., used to be the single biggest retail day of the year, officially kicking off the Holiday shopping season. In recent years, that one day has turned into five – some brick and mortar retailers are open on Thanksgiving Day, and the weekend sales stretch into Monday. And of course, online shopping is available anytime, day or night, from a desktop, pad or phone, with a concentration on Cyber Monday.
For those in charge of keeping the supporting data centers running and available through this unbelievable sales deluge, this can be a painful process. With online shopping predicted to once again break records this year, what can a data center manager do to mitigate the risk of failure? Look to these areas for answers:
With increased uptime demand comes increased anxiety. But knowledge breeds confidence. If a data center manager has the capability to track power, from the individual branch circuit, all the way up to the utility meter, surprises will be minimized. With the ability to monitor power usage and demand at that granularity, fluctuations will be noted immediately, and with the help of alerts, changes can be made before disaster strikes.
Optimizing power usage now - making sure that power is used in the most efficient manner possible - can give a data center manager that little bit of “wiggle room” that might be needed during high-traffic days. Case-in-point: power monitoring in real-time with a Data Center Infrastructure Management (DCIM) system can reveal “Zombie” servers – servers drawing power but not doing any processing. Removing these servers can provide the extra power needed to sustain service.
Power preparation is also crucial. A DCIM system will also help in testing and hardening a power chain against failure. Simulating a failure in a controlled environment will show the operator what would happen if a particular piece of equipment suddenly failed or was taken offline. This simulation will reveal vulnerabilities in the power chain and redundancy weaknesses that can be corrected before the big shopping weekend bursts into action.
Cooling is traditionally one of the biggest - if not the biggest - expense in the data center. Through real-time temperature monitoring throughout the facility, risk of failure can be minimized by uncovering hot spots so that load or equipment can be rearranged to avoid overheating.
A DCIM solution provides monitoring and mapping of data center cooling. You can use the DCIM’s rack, floor, and temperature mapping to move equipment into optimal locations controlling hotspots and raising ambient temperature. On average, for every degree ambient temperature can be increased, 4% savings can be realized on the utility bill. That is a substantial saving when considering that extra power needed to cover Cyber Monday activities.
Security is a huge consideration, with all the personal consumer information, credit card details, etc. that will be exchanged online. Data Center managers should take a good look at how secure their facility is from both a cyber and physical perspective. You can monitor for malicious and unintentional changes affecting your data center equipment, with a solid DCIM solution that would flag outdated versions of software and firmware currently running on your assets that may pose security vulnerabilities. This will address threats and provide time for risk mitigation before the big crunch on Cyber Monday.
It’s estimated that the average cost of a data center outage in 2016 was $740,357, or $8,851 per minute of lost revenue, with human error accounting for almost one-quarter of these unplanned outages. And that’s on an average day, not the busiest day of the year. Now is the time to test your data center for flaws in redundancy and disaster planning, not when online traffic doubles in a day. Whether it is a change in state or human error, a DCIM solution keeps constant track and can provide real-time information for equipment and floor plan optimization.
Scalability for online transaction processing is needed to handle the increased traffic to prevent crashes. Capacity is crucial for bandwidth and transaction speeds. DCIM can help determine what servers are operating at maximum capacity and which one are capable of scaling.
Edge computing speeds up transactions and cuts latency. More organizations are leveraging edge computing to bring the processing closer to consumers, but often have difficulty managing them due to a lack of visibility into operations. DCIM plays a critical role in helping managers gain the needed operational insights into the edge devices to keep performance levels and mobile shopping activities moving forward.
A good DCIM solution can help with all performance issues as mentioned above by offering data center managers a holistic view and the insight they need to plan for spikes in processing and associated power draw.
Opinions expressed in the article above do not necessarily reflect the opinions of Data Center Knowledge and Informa.
Industry Perspectives is a content channel at Data Center Knowledge highlighting thought leadership in the data center arena. See our guidelines and submission process for information on participating. View previously published Industry Perspectives in our Knowledge Library.