An unorganized data center is inefficient in many ways. Poorly placed equipment may not cool efficiently or receive...
By submitting your personal information, you agree that TechTarget and its partners may contact you regarding relevant content, products and special offers.
the right power, and could take up more space than it needs.
Unmanaged cabling, in particular, can cause data issues, especially if power and data run too close to each other. Power cables positioned too closely to data cables create heat, data-destroying electromagnetic forces and radio frequency (RF). These unmanaged cables can also stop the flow of cooling air in the under-floor plenum, or can even heat up that air, making it less capable of doing its job of cooling the IT equipment.
With lower efficiencies, greater unplanned downtime and higher costs, an unorganized data center is a significant burden on your resources. DCIM tools, data center cable management and more efficient zoning practices can help clean up that mess.
Start with a general cleanup
Carry out a full asset inventory of what you have, including servers, storage, network hubs and switches, as well as all ancillary equipment. This can be done with many different tools, but in a case like this, use a data center infrastructure management (DCIM) tool, such as those from Nlyte, Future Facilities or Emerson Network Power.
With DCIM tools, you get an active asset repository, along with the capability to draft out a full floor plan of where your equipment is now. From there, you can perform "what if?" analyses on possible layouts. For example, you could determine whether there would be sufficient cooling and power distribution if you move multiple racks to create a denser and space-effective layout.
DCIM systems also allow you to analyze and optimize your wiring layout within the constraints of what you have. You can also more easily identify what cables are no longer live, and begin cleanup with a lesser probability of any accidental downtime.
Don't overlook data center cable management
Using a DCIM system only takes you so far. A better approach is to go for fully structured cabling. There are two options -- under floor or above floor. With an under-floor approach, it is difficult to get to the cables, and tempting for an engineer to cut corners and not follow a structured cabling model. It is better to go for an above-floor, wiring tray-based approach.
Be sure to keep data and power cables as far apart as possible to avoid any electromagnetic force or RF interference. For easy identification, label and color cable runs. Cut each cable to the exact length required and don't loop them.
With a DCIM tool, you can receive automatic alerts related to data center cable management. For example, you can set alerts for when an engineer unplugs a cable to gain access to an asset and then plugs the cable back into the wrong port.
However, while this ensures more efficient data center cable management, it still leaves problems with equipment.
For a higher-cost option, consider zoning and tiering
Organizations can carry out zoning in many ways. The simplest option is to create zones within the data center that are dedicated to function -- for example, an area for servers, one for storage and one for distributed networking. In this way, IT teams can manage equipment and its associated cabling more effectively. However, in a world of increasing convergence, this approach is no longer that easy to follow.
Around 2005, there was a move to create highly specialized, zoned data centers. Here, organizations would build the facility with different areas serving specific functions, such as an entrance zone, a management zone and a data distribution zone. Although theoretically sound, it did not take off due to the large expense and constraints it placed on the need for a flexible IT platform.
Instead, the industry uses the tiering model when looking to build new data centers. The Telecommunications Industry Association published the definition of four tiers as reference architecture approaches to make it easier for people to understand the capabilities of each facility. The basic Tier 1 facility was just a room that held a server; Tier 4 was a mission-critical facility with multiple redundancy capabilities.
This has now been generally superseded by the Uptime Institute's tiering, which quantifies tiering by availability.
How to implement a data center cabling infrastructure
Avoid damage with good network cable management
Keep a data center cable patching project on track