Data Center Design Best Practices

Data centers play a critical role in the modern business sphere. With IT serving as an essential pre-requisite for business innovation; it’s quite common in fact, for a business’s capabilities to be built entirely around the competences of their IT department.

This change has brought a shift in infrastructural investment with increasing numbers of companies using data centers to create an advantage over their competitors, while at the same time increasing efficiencies and reducing operating costs.

However, data center energy consumption has mushroomed in the last 10 years with each consuming up to 100-200 times as much energy as an office. To put this into perspective, that accounts for 3% of global electricity use. In order to be flexible, efficient and reliable, data center design has to adapt to the rapid pace of demand and technology.

Here are some best practices for data center design to ensure your business is able to cope with the growing complexities of data center management.

Performance

Data center infrastructure has a direct impact on network performance. When high density configurations, virtualization and other capacity-boosting methods were adopted, the necessity for a robust business-critical application infrastructure was overlooked. 

To successfully maximize network performance, there are three parts of infrastructure that must be considered; racks and cabinets, structured cabling and cable management. Here are 5 performance considerations to take into account for data center design:

1. Ensure your solution supports both copper and fiber media.
2. Maximize channel performance by using a cabling solution that has co-engineered cable and connectivity.
3. Choose flexible and scalable racks and cabinets that can accommodate higher weight thresholds.
4. For improved cable support, protection and airflow, incorporate adjustable rails and wider vertical managers, along with integrated cable and airflow management options.
5. Choose a manufacturer that is up to date with current standards as manufacturing companies are aware of new, upcoming tech requirements long before they’re published.

Adaptability

Up until now, long term adaptability was accomplished by oversizing the infrastructure system and letting the data center grow into its infrastructure over time. However, this is neither capital or energy efficient. Consequently, infrastructure systems must be designed with greater flexibility and scalability in mind, allowing for the optimal sizing of a data center.

Because data center space is a precious commodity, infrastructure systems should always be built with maximum optimization in mind. Nevertheless, design elements evolve with the ‘Internet of Everything’ and the related demands of hyper-connectivity. As such, future adaptability is unpredictable and difficult for companies to plan for. 

However, organizations can do the following to optimize their data center space in the interim:

• Select racks or cabinets as your basic building block for data center density.
• Pick racks and cabinets with higher weight limits, sufficient depth and height to support vertical growth.
• Select a cable management system that can support existing cable density, provide ample room for growth, mitigate airflow restrictions and support both copper and fiber cables.
• Choose connectivity that supports high density and mixed media.
• Use cable with smaller Outside Density (OD).
• Consider patching outside the rack and cabinet (e.g. overhead) to conserve space for equipment.
• Select a rack or cabinet solution that easily integrates with overhead pathways.

Time

Data centers are becoming larger and more complex, yet the need for fast setup is growing in intensity as a result of consumer demands. Data centers must be able to change with business requirements resulting in faster deployments, reduced cost, and allow for easier moves, adds and changes. 

With 90% of active equipment being replaced within a 5 year period, it’s paramount that DC infrastructure is able to handle regular updates; supporting cable upgrades (such as replacing copper with fiber), increasing cable and port densities, as well as the essential maintenance of Top Of Wrack (ToR), Middle of Row (MoR) and End of Row (EoR).

Modular solutions pave the way for flexible and scalable building infrastructures, combining the advantages of standardization and customization to deliver efficiency. Data center modularity allows for easy expansion and offers the following time-saving qualities:

• Choose modular racks and cabinets that assemble quickly, have adjustable rails, higher weight thresholds, and taller heights to easily accommodate new equipment and system modifications.
• Select modular options for optimizing airflow management in a rack or cabinet and efficient cooling solutions (with a >5kW density).
• Pick cable management that is easy to install, change and available where needed. For example in the front or back of the rack/cabinet.
• Opt for wide vertical cable managers that support changing cabling needs, minimizing obstruction from the cables to support better airflow, maintenance and adaption.
• Choose integrated pathways for rack/cabinet optimization.
• Select pre-terminated solutions that allow for quick installation, easy moves, adds and changes, and easier migration to newer technologies.
• Pick a modular solution designed to optimize installation and deployment time which will also support future network changes, computing power and technology upgrades facilitating growth.

Read more on the benefits of data center modularity.

Experience

As outsourcing grows in popularity, it’s integral that the correct partner is chosen to collaborate on a data center project, to provide the experience, flexibility and scalability capable of facilitating tomorrow’s tech innovations. Additional customization is also an important consideration to include in data center planning, with design factors needing to accommodate flexibility as well as specific modular needs. 

During the installation phase, it is critical that theory and product work together seamlessly. The data solution must be easy to install, quick to deploy and easy to manage. Logistics must also be coordinated to guarantee timely delivery.

Choice of manufacturer is a crucial component in data center success. Your selection of manufacturer should be made in accordance with their ability to provide resources to help coordinate your project through the entire design and installation process, creating additional opportunities to standardize and support extended location capabilities.

In addition, manufacturers should have across the board expertise including new cooling technologies such as close-coupled cooling, power, connectivity, cabling, physical infrastructure, cable management, and pathways to guarantee flawless operation. This expertise should be utilized to extend equipment life, reduce cost and address the unique challenges specific to data centers, accomplished through a singular point of contact and ongoing support.

Sustainability

Although the definition of sustainability can have multiple connotations in relation to data center best practices, it is often associated with improving processes to avoid exhausting or destroying natural resources. The perceived importance of sustainability in data centers varies, but is growing in prominence as environmental issues begin to take precedent on the organizational agenda. This transition can be simplified further by working with a sustainable manufacturer who can develop solutions that reduce the impact of operations on the environment. A manufacturer that has ISO 14001 certification means that it is actively working to reduce the environmental impact of their processes, products and services. This may materialize into projects that incorporate green energy and design as well as a reduction in on site waste and installation.

With data centers consuming up to 200 times as much electricity as standard office spaces, (a figure set to double every four years); the design and best practices of data centers will play an increasingly important role in the reduction of energy consumption and ongoing technological sustainability. As discussed above, maximum efficiency can be achieved by considering factors such as performance, time, space, experience and sustainable practices.

Although data center design is more complicated than ever, prioritizing factors mentioned in this blog will create an efficient, reliable data center that optimizes operations and environmental feasibility wherever possible.