A network cable management guide

Best practices for network cable management

Network engineers should consider cabling best practices to simplify management once installed. Following these steps helps with network maintenance, troubleshooting and simplifies the process of implementing infrastructure upgrades in the future.

1. Corral the cables

In a data center, keep cables out of the traffic path -- from floor level up to head height -- and out of the workspaces behind racked equipment. Use cable trays overhead or under a raised floor to get from row to row or from one rack to a non-adjacent one. For adjacent racks, use cable supports and organizers on the sides and backs of racks.

At the equipment end, whether it's a server or switch, leave a little slack in the patch cables to easily plug and unplug from the equipment. The same advice goes for patch panels.

Outside data centers, secure cables in closets, passages through walls and above drop ceilings using metal straps, plastic zip-ties or Velcro. Cable races, trays or conduits are also helpful for keeping the cables from snagging on other infrastructure and vice versa. This makes it easier to trace a cable through the physical environment for troubleshooting and repairs.

2. Mind the weight

Bundles of network cables can be quite heavy. If cables aren't lying on the floor, always provide adequate support that can double as cable organizers in some cases. Supply support with a significant safety margin -- more than necessary for the current cable set -- in anticipation of adding more or heavier cables.

Weight can be a pressing concern when upgrading older cables. Modern cables that support Power over Ethernet (PoE) are often significantly heavier than the older cables they replace. This can be a problem with overhead raceways and cable supports in data centers, wiring closets or above raised ceilings.

3. Differentiate with colors and labels

Every cable has two ends, and it's critical to handle the same cable at both ends. To simplify this, label both ends with a number or an alphanumeric code that embeds information about the run, such as the closet it originates in. Cable codes don't have to be unique overall, but they should be unique with respect to the places where the run ends, such as a closet or the data center.

It's often not useful to label cable ends with anything more specific than a closet identifier in a cable labeling code. Some network pros want to embed hardware or port identifiers, but these components are subject to change and would require relabeling. It's better to use a component that is less likely to change, such as the wiring closet where a cable ends.

Color coding is another way for network staff to easily identify a cable during troubleshooting or moves, adds and changes, especially in a data center. Cables for data connections can be one color, storage network connections another and KVM connections a third.

The drawback to color coding is it's harder to freely reuse a patch cable. To avoid that, use colored tape near the cable ends instead of colored cables. Pair labeling and color-coding by printing the cable code on a colored label.

4. Document everything

It's hard to manage cables if their locations are unknown. Regardless of the labeling and color-coding system a team settles on, they should clearly spell out and maintain the standards in central, shared spaces with other organizational standards and practices documentation.

In areas where cables run out of sight, such as under a raised floor, the team should also ensure they have maps or diagrams that clearly show where the cables run. It's sometimes difficult to find cables or see much under full racks of equipment, so it's important to have documentation that references labels and colors. This is also relevant for when cables run through a conduit between floors or in a buried conduit between buildings.

To ensure documentation stays current, both IT and facilities teams should include updating cabling maps and documentation in their change management, construction and renovation processes.

Cable management practices to avoid

Bad cable management practices are the inverse of best practices. Network professionals should avoid the following when managing network cables:

• Lack of cable use case knowledge. When network professionals don't understand the use cases for a specific type of cable, they struggle to identify which cables are most appropriate for different situations.
• Improper cable routing. Improper cable placement or routing can complicate cable management and create safety hazards. Don't run cables loosely from row to row or rack to rack, as this can add stress to the cables and lead to malfunctions. Sharply bent cables can also fail. Don't run cables across the fronts and backs of equipment in racks. Instead, use the cable management races on the sides.
• Failure to follow plans. Having a plan for labels and color-coding is not enough; network professionals must also follow it.
• Outdated documentation. Network professionals must update documentation during moves, adds, changes, construction and renovation.
• Crowded trays or conduits. It can be difficult to access cables in trays and conduits filled beyond their capacity. In addition, cable trays that aren't supported well enough might not stay in place under load.
• Inadequate cable placement. Not tracking where cables terminate or originate in long conduits can create network management and maintenance problems. Ensure cables aren't bent -- which can eventually malfunction -- or coiled around sources with electromagnetic interference, such as elevator motors.

Cable management optimization strategies

In addition to following best practices, network teams should also focus on other strategies that enhance their approaches to cable management. For example, rightsizing, along with routine spot checks and audits, can help network teams proactively manage, monitor and administer their networks.

Rightsizing

While cable specifications can indicate the maximum length of a particular type of cable, they can't show the optimal length for a network's specific requirements. Rightsizing is the process of gathering data and adjusting infrastructure components based on expected needs It helps admins identify performance requirements, plan for future scalability efforts and upgrades, and account for costs.

A successful network design, influenced by a proper rightsizing approach, strikes the right balance between conserving materials and flexibility. For example, organizing cables to be only as long as needed enables network teams to have enough room to make minor changes to the network over time.

Examples of rightsizing practices include the following:

• Relocate one end of a cable to make room for another rack.
• Position cables from one side of a wall to another.
• Route the middle of a cable to make way for new installations, such as pipes, ventilation conduits or sprinkler systems.

When network teams install cables, it's important that they leave some slack for cable repositioning. This prevents them from having to re-pull a cable or insert a switch or repeater into the run.

Spot checks and audits

It's great if network teams have all the documentation about their networks, especially when required updates are built into standard processes. However, to keep documentation accurate and useful, network teams should regularly audit the cabling and documentation together.

Pick a space and ensure the cabling is up to standards and the documentation is accurate. This can be a separate maintenance process or piggybacked on other trouble tickets. Although audits should be regular, they can be random or based on a plan and schedule.

Consider wireless connectivity

Organizations that want to reduce the number and weight of cables can consider switching their networks to wireless endpoints outside the data center. This option might be especially useful in endpoint-dense locations such as open-plan office floors and classrooms.

However, cabling is still necessary in wireless networks. Wireless endpoints still need connectivity and PoE. Other endpoints, such as big, shared printers, still remain wired.

John Burke is CTO and a research analyst at Nemertes Research. Burke joined Nemertes in 2005 with nearly two decades of technology experience. He has worked at all levels of IT, including as an end-user support specialist, programmer, system administrator, database specialist, network administrator, network architect and systems architect.