|
According to the Association for Computer Operations Management (AFCOM), one out of every four data centres will experience a business disruption within the next five years that’s serious enough to affect an entire company’s ability to continue business-as-usual. For 40% of them, it will take longer than a day to get records back online, and it will take more than a week for a full system resolution for 7% of them. When the problem behind the system outage is finally uncovered, there’s a 21% chance it was related to the cabling.
Even in situations where an IT worker selects good components and tries to develop a systematic approach for building structure into the cable plant, the situation will grow out of control when he hasn’t planned for adequate cable management or lacks the discipline to manage the physical environment. Imagine trying to locate and troubleshoot, or move, a cable in the entanglement shown in the photo. History and many hundreds of visits to data centres have led me to believe that this type of loss is based on poor design, installation and product choices.Choosing a successful cable methodologyLet’s start with design. There are basically two methodologies for building cable infrastructure: point-to-point, where server HBAs attach directly to switch/director or device ports, or, structured, where server, director/switch and devices ports all connect to passive elements for centralized management control.
Problem with point-to-pointThe major difference between the two methods is the ability to manage change — planned or otherwise. The direct point-to-point connection method has fiber optic jumper cables running under the raised floor between the active equipment. The exact routes of the cables are usually not well documented, and only the end points are known, based on some sort of label on each end of the cable. The cabling usually becomes a tangled mess under the floor, which makes rerouting cables difficult. When system cabling changes are made with the point-to-point connection method, system equipment doors must be opened, service professionals have to unplug/plug FO jumper cables inside the active equipment, and cables must be pulled back and rerouted under the floor. All this activity puts system connections and equipment at risk. When a site is not designed for the future—and the cabling not installed properly—the result can be chaotic, as loss is not only based on the design but on the installation. Conveying systemA conveyance system, which lays out precise cable pathways, serves several valuable installation purposes with regard to cabling and, without it, there can be high losses. The first step in troubleshooting a cabling issue is locating the cable to be evaluated. Defining precise cable pathways through the use of a conveyance system and documenting the cables that traverse those pathways is an effective way to help locate cables in mid-span.In addition to helping locate specific cables, a pathway will prevent haphazard routing. As cables are routed under the floor where they overlap and become entangled, they experience unnecessary stress. And while they may continue to function properly, this stress may cause micro-fractures or other results detrimental to the infrastructure’s performance over time. After creating a precise pathway, the conveyance system brings the FO cables up off of the concrete floor and separates them from copper, power and other cables and fixtures (not to mention possible flooding or gnawing rodents), protecting them from damage. It also places them in a protected environment where they won’t be trampled underfoot and crushed as plumbers, electricians and other subs work under the floor. Cable management recommendationsThe best way to implement data centre fiber optic cabling is via the structured cabling system. This method uses multi-fiber trunk cables and FO patch panels to establish defined cable pathways and designated cabling connection change areas in the physical cabling design. As soon as these cable pathways and designated connection change areas are established, the FO cabling for system hardware modifications can be easily isolated in the infrastructure. This greatly simplifies cabling modifications by ushering change activity to defined locations rather than disconnecting, moving and reconnecting discrete FO jumper cables scattered under the raised floor. Five major elements of structured FO enterprise cablingIt is important to address five key areas that affect the performance, reliability and manageability of structured fiber optic enterprise cabling: cable assemblies, patch panels, distribution cabinets/racks, accessories and conveyance systems. Care must be taken when selecting or developing any of these elements. It’s important to create a structured cabling environment based on a central patching location (CPL/MDA); also, consider moving to MTP trunking solutions to best manage the density and weight issues associated with these cables. Thinking long-term solutionIt’s also crucial to keep in mind both the small details and the big picture when developing a cabling strategy plan. Remember, short-term solutions can frequently address an organization’s immediate concerns and needs but might, in the long run, contribute to greater disruption and exposures. The key is to plan strategically and institute processes that optimize risk management and accommodate future growth. Rick Dallmann is marquee accounts manager for CXtec, a provider of new and certified pre-owned computer networking and technology equipment. He can be reached at
This e-mail address is being protected from spambots. You need JavaScript enabled to view it
. Reprinted with full permission of Network & Cabling Magazine www.networkcablingmag.ca
|
News & Issues 
