Data centers are a hot topic these days, especially when it comes to energy efficiency and maximizing capacity in the data center. One area that is often overlooked is the setup of the data center floor. Knowledge Center contributor Rick Sawyer offers data center administrators seven tips to follow to keep their data center floors performing at optimal levels.
When approaching an energy efficiency program, it is important to keep in mind the day-to-day factors that data center administrators face. They must ensure that current facilities and equipment are working optimally in their existing settings. Some procedures are obvious, such as keeping data center aisles clear so that air flow is optimized. But some measures are not so obvious and can cause problems if improperly addressed.
One area that is often overlooked-but hugely important in maintaining a healthy data center-is the setup of the data center floor. A well-maintained and raised floor will ensure equipment stability and protection for servers from power fluctuations. A poorly maintained floor can hinder a cooling system's performance and put equipment in jeopardy. Here are seven tips that data center administrators can follow to keep their floors performing at optimal levels:
Tip No. 1: Set up floor tiles properly
Perforated floor tiles should be opened in cold aisles only-and closed in the hot aisles. This allows for optimal circulation patterns and ensures that servers are adequately cooled.
Tip No. 2: Reduce air leakage from floor cable cut-outs
In raised floors, cable cut-outs that are not properly sealed can result in leakage, which reduces the data center administrator's ability to keep the room properly cooled. Standard methods to seal cable cut-outs include fastening a brush fixture around the hole or using a fiberglass seal.
Tip No. 3: Use cleaning equipment properly
Due to the highly sensitive nature of the equipment in a data center, a standard vacuum cleaner is not suitable for use. A vacuum that has a high efficiency particulate air (HEPA) filter can remove at least 99.97 percent of airborne particles that are 0.3 microns in diameter. Particles of this size are the most difficult to filter and are considered the most penetrating particle size. Standard vacuum cleaners will allow particles to circulate and settle on equipment, which can lead to damage.
Tip No. 4: Ensure the floor support system is intact
Server racks get moved around frequently. So, when equipment is being relocated, it is important to make sure that the floor tiles are properly reinforced underneath. This can be typically done with steel reinforcements.
Tip No. 5: Prevent a domino effect if a floor collapses
When a data center floor is not properly supported (due to an improper support system), the integrity of the floor is compromised. One weak area can spread to other areas, creating a domino effect. Looks can be deceiving; a blade server can weigh hundreds of pounds.
Tip No. 6: Be grounded
Data centers have a multitude of electrical and magnetic fields. Because of this, equipment protection is a key consideration. The data center floor should be grounded. If it is grounded, this will, in turn, ensure that the servers that are sitting on the raised floors are also grounded. Not grounding a floor or the equipment on it can lead to server processing errors.
Tip No. 7: Keep the floor clear
It is common for data center staffs to use rolling carts when maintaining or updating equipment. However, most raised-floor manufacturers don't account for this added weight for sustained periods of time. Rolling carts and other obstructions also hinder the cooling efficiencies within the data center, so it is best to keep aisles clear and items off the floor whenever possible.
Energy efficiency in the data center truly starts from the ground up. By establishing and following some standard rules on floor maintenance, data center administrators can help the facility run more efficiently. This ultimately prolongs the life of the floor as well as the equipment that sits on top of it.
Rick Sawyer serves as senior principal, Critical Facilities Assurance, at HP Critical Facilities Services. Rick's expertise includes project design, equipment specification and selection, contracting, testing, performance measurement and national engineering standards for global data center development. In addition to the design and construction of major, mission-critical facilities in the United States, Rick has developed data centers in Japan, Brazil and Argentina.
Rick is also a frequent presenter at Data Center World Conferences and a board member of AFCOM's Data Center Institute. He can be reached at RSawyer@hp.com.