0
Share 
Massive data is created every day and stored in data centers, and the phenomenon is only poised to grow. IT will be spending 237 billion dollars on data centers in 2021 worldwide as per projections.
So naturally, for today’s data centers, there will not be a shortage of competitors, and hence they cannot afford to make mistakes or lose capital.
The core success behind a data center lies in its Design Efficiency and Performance, and Design Expandability.
Data center design best practices determine where the data center stands in the market, and how good an asset it can be for your business.
To help you stay on top of your game, we have listed some of the best data center design practices, which are guaranteed to help you in the long haul.
Irrespective of the size of a data center, these Physical Elements will inevitably pop up in its design. Therefore, neglecting to adopt these practices can directly impact the success of your data center.
The primary power source to most data centers is the more comprehensive municipal electric grid, which passes through one or more transformers to ensure the correct voltage and type of current (AC or DC) flows to the entire infrastructure.
Power backup systems like UPS and Backup Generators are used to ensure uninterrupted data delivery and reduce or eliminate downtime.
Sometimes, a single UPS is sufficient for a data center.
But a tier 3 or higher class data center might require multiple UPS units to keep systems online until the backup generator fires up. UPS units usually sustain servers for about 20 minutes on battery power.
In a data center, IT equipment, cooling equipment or HVAC, lighting, and other miscellaneous infrastructure use power to run. Consequently, a fortune is spent on IT and HVAC equipment that utilizes the most energy in the data center. This makes sense, as these pieces of equipment are active 24/7and produce immense heat.
Cooling tech is designed considering several factors in a data center. Mechartes, by studying and analyzing the conditions within a facility, offers data center validation services, which account for energy efficiency and optimizes costs.
You must design the power and cooling solution to adapt accordingly, as a data center is evolving.
This setup often leads to additional and extensive capital investment, and you can avoid these if the solutions are sustainable.
Sustainable design practices or Green Data centers have proven efficient and can easily accommodate a growing facility if implemented and maintained correctly.
According to studies conducted by The Independent United States Data Center Energy Usage Report in 2016, data centers were predicted to consume 73 billion kWh in 2020.
“A footprint is the negative impact of a sector on the environment. A handprint is the positive environmental impact: how a sector can reduce the footprint of other economic sectors and society as a whole.” – Alliance to Save Energy.
So, how does a green data center work?
A Green Data center – or sustainable solution to meet growing energy and cooling requirements – is designed for maximum energy efficiency and minimum environmental impact.
The entire facility is designed sustainably or run by renewable natural resources like solar, wind, geothermal, or hydroelectric energy. As a result, they considerably improve energy and cooling efficiencies.
Investments for a sustainable solution are expensive upfront and require proper operations and maintenance. But they also serve as a long-term cost-saving solution.
You can say how energy efficient a data center is with Power Usage Effectiveness (PUE) and Leadership in Energy and Effectiveness Certification (LEED) ratings.
PUE rating greater than 3.0 indicates inefficiency, and 1.0 means high efficiency in converting energy.
Apart from Power and Cooling, a green or any high-performing, competitive data center also practices sustainable measures about floor space, cabling, and other more minor details like rack design.
Data centers with cabling support common topologies (like ToR, MoR, and EoR), and new configurations like leaf-spine offer flexibility and scalability.
In a data center, 90% of active equipment can be replaced. To meet emerging needs, adopting modular cabling solutions that accommodate growth without compromising performance will save time and capital in the future.
In addition, using cable organizers and color-coding eliminates clutter that hinders airflow, impacts cooling efficiencies, and makes overall maintenance easier.
When designing racks and cabinets, using filler panels allows airflow and facilitates expansions for the future. Racks and cabinets supporting heavy equipment, high-density server racks, and good depth-height ratios for vertical expansion can also aid expansion.
If possible, select manufacturers with ISO 14001 certification who actively work towards reducing the impact on the environment through their processes, products, and services.
Data centers hold a valuable asset and a security breach diminishes the profile and credibility of the facility. Security features like CCTV surveillance, bio-scanners, multi-factor authorization, workforce, and restricted access to personnel can help monitor movements and restrict unwanted access to the facility.
Data centers like Switch have multi-layer security to restrict physical access to the facility, starting from fencing, bollards, and mantraps.
To ensure a data center operates at 100% efficiency and stays competitive, monitoring and managing operations, infrastructure, and all other tasks within the facility are vital. Therefore, non-physical elements play a critical role.
Software Solutions like Data center Infrastructure Solutions (DCIM) keep track of all activities and changes made in a data center facility and update them in real-time.
This software eliminates manual input or spreadsheet databases, reducing inaccuracy, and enables visualizations, reporting, and analyzing.
“It is difficult to achieve the most advanced levels of data center effectiveness without extensive use of DCIM with workflow.” – Andy Lawrence, 451 Groups.
Integrating DCIM with other systems like Building Management Systems (BMS) and IT System Management (ITSM) solutions improves workflow.
When workflow systems are consistent and are integrated, all tasks become traceable, transparent, and auditable. In addition, as tasks are allocated and tracked, it provides documentation and gives a centralized view of all operations.
This optimizes resources, time, and equipment while improving the efficiency and productivity of both staff and the data center.
Management Solutions can also be used to monitor the Energy Consumption and Efficiency of the facility. As a result, they eliminate the risk of outages by updating statuses in real-time, monitor cooling infrastructure, and help detect potential issues and eliminate them with reduced capital risks.
Rackspace has been managing data centers and colocation environments since 1999 and decided to go 100% green in 2016, predicted to save millions of dollars for 20 years.
The Sussex facility alone holds 130 MW with a PUE rating of 1.15, saving $60 per kW per month. They use green generators and power supplies assembled in workshops offsite and installed as finished units.
The location itself is “hard to locate by common man,” further protected with primary fencing and biometric scanners to restrict access. The tech ops room controls and monitors all operations within the facility using integrated software.
The 50,000 servers in the Rackspace Data center are assembled offsite and installed to avoid additional dust and pollutants. In addition, it uses indirect air cooling systems to maintain optimum hot aisles temperatures of 32° – 34°C.
Two hundred thousand meters of cables are color-coded for easy maintenance. By using transparent-pane LED lighting throughout the site, they save 12 months of energy costs.
Sun pipes are also used for illumination, as bright as the LEDs in the right conditions. The roof is designed to harvest as much rainwater as possible used in the cooling units.
Ten units of power supply in N+2 configuration are used for each 3MW data hall and are run at lower energy levels. If one unit fails, the other units take up more load to prevent downtime. They have considered expansion for their site and have made accommodations for the same.
To summarize, study your site inside-out while determining power and cooling infrastructure and go green wherever possible. Then, plan for the future and avoid wasting time and capital on expansion. To stay competitive, adopt the data center best practices in finer details too.
With a team of over 50 highly qualified engineers with multidisciplinary expertise, Mechartes has enabled several businesses with their data center validation services to provide optimized and performance-based designs using CFD & FEA simulations. Make the best use of their services to implement the best data center design practices for your business.
0
Share 
facebook