6 Best Practices to Improve Your Data Center Operations
There are several best practices CTOs can follow to ensure their IT operation is efficient, running within capacity, and executing as productively as possible.
Join the DZone community and get the full member experience.Join For Free
In his article, ROI Valuation, The IT Productivity GAP, Erik Brynjolfsson states, “The critical question facing IT managers today is not, ‘Does IT pay off?’ but rather, ‘How can we best use computers?’” This is not a simple question for CTOs to answer because each data center and IT operation is unique, with a multitude of variables affecting the overall operation. Two different companies with almost identical IT ecosystems yet one might have a fraction of their competitor's productivity, argues Brynolfsson. However, there are several best practices that CTOs can follow to ensure their IT operation is efficient, running within capacity, and executing as productively as possible.
1. Clean Up and Declutter
“Cleanliness is godliness” as the old saying goes, and it could also be stress-relieving when it comes to IT. Servers and networking equipment all have set lifespans and old equipment should be decommissioned on a schedule defined by the manufacturers. Old equipment should be properly destroyed, recycled, or returned to the manufacturer, with all data wiped clean to ensure proper security.
2. Measure PUE
Google is one of the largest cloud providers around and they run massive data centers, some containing thousands of servers. Needless to say, they know a thing or two about keeping data centers operating at their peak. Google argues you can't manage what you don’t measure, and the search engine leaders characterize its data center's efficiency performance by measuring energy use.
"We use a ratio called PUE - Power Usage Effectiveness - to help us reduce energy used for non-computing, like cooling and power distribution," says Google. They measure samples at least once per second. They also take into account the weather as seasonal variations have a notable effect on PUE.
3. Manage Airflow
"Good airflow management is fundamental to efficient data center operation," notes Google. Hot spots should be completely eliminated, while blank pages or filler panels should be used for empty rack slots. This helps cut down on dust issues. Filler panels ensure airflow is undisturbed and dust can be blown away easily. Proper airflow is critical for cooling.
With properly installed filler panels, the air blows where it is supposed to, which keeps the rack at the proper temperature while the server is operational. This is one of those minimal cost investment-maximum performance enhancement solutions that are well worth the time required to install filler panels.
4. Monitor Everything
According to Plant & Works Engineering man is “The best condition monitoring device ever invented,” but his position is under threat by so many of today's monitoring tools. An experienced IT technician might understand every nuance of a system he has been working with for years is, but today’s process and monitoring tools can go much further in understanding an operation's processes than a man can. Business process management (BPM) software, robotic process automation (RPA), and AIOps all evolved from early generation IT Operations Management tools and aim to enhance IT operations with automated processes as well as potentially self-heal a system.
Gartner defines AIOps as a platform that utilizes big data and AI to enhance IT operations functions like monitoring, automation, and service desk activity with proactive, personal, and dynamic insight. "AIOps platforms enable the concurrent use of multiple data sources, data collection methods, analytical (real-time and deep) technologies, and presentation technologies," says Gartner. An AIOps solution can learn and monitor a company’s day-to-day IT operations, analyze the entire IT system, then either fix issues or alert those who can.
5. Capacity Planning
Capacity planning is the process of determining an organization's system and operational needs in an attempt to understand and meet the changing IT and energy demands due to the sale and use of a company's products and/or services. Capacity management attempts to counterbalance the right number of users with the right performance at peak usage to ensure an enjoyable end-user experience. IT capacity planning involves estimating the resources required for a company's storage, hardware, software, and connection infrastructure so the system works as optimally as possible, while also limiting wasted capacity. Capacity management looks to add or subtract CPUs, memory, and storage to a physical or virtual server.
CTOs should optimize their systems while continuously reviewing their cloud usage at the application level by correlating business demand with cloud service utilization. CTOs should plan for growth and predict upcoming costs with advanced analytics. By monitoring everything, CTOs will have a great understanding of CPUs, memory, storage, and power usage for the IT estate and act accordingly.
6. Take It to the Bank
Data centers are the center of the universe for most companies these days. Increasing capacity, reducing redundancy, and improving overall efficiency should be the goal for every CTO. In many ways, CTOs are being asked to do the impossible. Big Data is only getting bigger and more unwieldy. The five Vs of Big Data have expanded to seven – Volume, Velocity, Variety, Variability, Veracity, Visualization, and Value.
Never before have so many companies had so many ways to collect, track, quantify, and even visualize data, but it's almost an overwhelming undertaking. The 5G rollout and the expansion of the IoT are only going to make data collection, integration, virtualization, and even visualization more difficult.
“In the information economy, the scarce resource is not information, but the capacity of humans to process that information,” warns Brynjolfsson. Many of the best practices above remove humans from the equation, which isn't a bad thing, as our time can usually be better spent on higher-order and creative endeavors rather than repetitive, mind-numbing undertakings.
“Too often, the flow of information speeds up dramatically in highly automated parts of the value chain only to hit logjams elsewhere, particularly where humans must get involved and processes aren't updated. The result is little or no change in overall performance,” says Brynjolfsson. The time to get one's data house in order is now and perhaps that's with a little more automation and a little less man.
Opinions expressed by DZone contributors are their own.