Wednesday, December 19, 2012

Structured Network Cabling Primer (Part 3 of 3)


Structured Network Cabling Primer

Part 3 of 3

As mentioned earlier in this series, when working with cloud-based applications, such as Google products, its helpful to know networking basics to communicate effectively with technicians who provide the structure that make cloud use possible. After exploring the office environment in Part 1, Part 2 covered and the wiring closet environment, including patch cables, horizontal runs, vertical risers, and the communications equipment that allow devices on the same network segment to communicate with each-other. This edition, the last in the series, focused on the equipment room and how it all ties together.

The equipment room, or server room, is like a wiring closet, but bigger and usually more complicated.



A more advanced example of a network configuration
A more advanced example of a network configuration

Equipment Room Environment

Equipment rooms usually have patch panels that terminate the vertical riser cabling from wiring closets.  Patch cables are then used to connect the patch panels to hubs or switches, again, just like in wiring closets.  Additionally, equipment rooms usually have one or more routers or gateways that connect different network segments together; that is, they act like a gateway to a network segment and they route data to and from that network segment.

Router / gateway devices can often perform complex tasks with the data they handle.  For example, they can act like a firewall to prevent certain types of traffic from entering or leaving the network segment or only allow certain traffic to or from specific hosts. They can provide a safe, private network space and translate network traffic from the private network to another network (often called "Network Address Translation" (NAT) or "IP Masquerading").

Yet another device that can be found in an equipment room is a Modem.  Modems are used to MOdulate / DEModulate signals of one type (such as an Ethernet network) to another type -- such as cable (e.g., coaxial cable from a cable provider), ADSL, SDSL, ISDN, Optical Carrier lines (e.g., OC-48s), T-Carrier / Digital Siglan lines (e.g., T1s, DS3s, etc.), ATM, etc. -- that usually belong to the Internet Service Provider (ISP).

An important concept to understand is the point of demarcation, otherwise known as the demarc.  The demarc is a transition point in the network; on one side of the demarc is your equipment -- the stuff you're responsible for maintaining; on the other side of the demarc is the equipment your Internet Service Provider (ISP) is responsible for maintaining.  Problems that occur on your side of the demarc are yours to resolve; problems on the other side of the demarc are for the ISP to resolve.

Knowing exactly where the demarc is is extremely important.  Sometimes an ISP will provide or lease a modem to an organization and support the modem and all of it's problems.  In that case, the demarc could be the port on the modem that you patch into your router or gateway.  Sometimes you have to provide your own modem and your ISP only provides a wire dangling out of a conduit.  Knowing exactly where the demarc is located is critical for when problems crop up (and they will).

Labeling Cables

All cables need to be labeled.  The ANSI/EIA/TIA-606-A Standard asserts that identifiers (i.e., the string of characters identifying a telecommunications component) as:
"An identifier is used in labeling telecommunications infrastructure components such as cable, racks, telecommunications rooms, equipment rooms, pathways, and telecommunications outlets.  
It is a unique set of numbers, letters, or a combination of both sets that are not repeated within the administration of the system.
Identifiers shall be:
  • marked at the end of the element to be administered
  • unique (e.g.,  no cable identifier should be the same as a pathway identifier)"
Panduit Corporation has developed a scheme for labeling UTP (Unshielded Twisted Pair) cables that is designed to provide as much information about each cable on the cable itself.  As a result, just by looking at the cable, one can find the patch panels a particular horizontal run or vertical riser terminates.

Each label consists of two parts -- where the cable starts and where it ends separated by a colon.  Therefore, in order to build a label, a consistent naming scheme -- such as the scheme developed by Panduit Corporation -- needs to be implemented that clearly, uniquely, and unambiguously represents cable termination points.

For termination points at patch panels, the Pandiuit scheme lists the location of the closet (e.g., the room number), the rack of communications gear that contains the patch panel, the row of the rack for the particular patch panel, and finally, the position or port on the patch panel.  Consider the following example -- a cable terminates in Closet 01A, Rack B, Row 2, position 4:

An example of part of a rack in a wiring closet demonstrating identifiers using the Panduit labeling scheme
An example of part of a rack in a wiring closet demonstrating identifiers using the Panduit labeling scheme

Cables that terminate at wall outlets follow a similar scheme; they list the location (e.g., the zone), the outlet, and the position or port on the outlet.  The following example specifies Zone D, outlet 05, position 03.

An example of  a work outlet demonstrating identifiers using the Panduit labeling scheme
An example of  a work outlet demonstrating identifiers using the Panduit labeling scheme

Consequently, if a horizontal run connecting the patch panel and outlet in the above examples, the label for that cable would be:

01A - B02 - 04 : D05 - 03

If, instead of labeling a horizontal run connecting a port in a patch panel with a port in a work outlet, a vertical riser connecting two patch panels in two equipment rooms needed a label, use the patch panel scheme to represent both patch panel locations.  For example, if a riser connected a patch panel port in room 02A, rack C, row 07, position 21 with a patch panel port in room 01A, rack A, row 08, position 32, the label would read:

02A - C07 - 21 : 01A - A08 - 32

In addition to labels, colors are often used to identify the purpose of a port or cable.  Some installations use color identifiers on the cable, some use color of the label identifying the cable, and some installations try to use cable jacket colors.

The standard colors are specified in ANSI/EIA/TIA-606-A.  This table lists the colors along with their Pantone numbers (used to communicate specific colors composed of 13 base pigments; developed by Pantone, Inc.):

ColorPantone ColorPurpose
* RedPantone 184cKey telephone systems
* OrangePantone 150cDemarcation points
* YellowPantone 101cAuxiliary circuits (e.g., alarm systems, maintenance, security systems, etc.)
* GreenPantone 353cCommunications equipment on the customer's side of the Demarc
* BluePantone 291cHorizontal connections (e.g., work area outlets and network stations; HC connections)
* PurplePantone 519cCommon equipment (computers, printers, multiplexers, etc.)
* WhitePantone 264cFirst level cross-connects (e.g., main equipment room to telecommunications closets; MC connections)
* GrayPantone 422cSecond level cross-connects (e.g., between telecommunications closets; IC connections)
* BrownPantone 465cInter-building cross-connects

Other Important Points

Here are a few other important points to consider when designing and building a network.
  • Try to use components that are of the same cable categorization; for example, use Cat6 cables with Cat6 plugs; otherwise, the nework will only perform up to the point of the lowest categorization component.  So, if you use Cat6 cables with Cat5e plugs, you will only get Cat5e performance.
  • Use wire termination and punchdown techniques consistent with the specifications for the cable you're working with; if you're using Cat5 or Cat5e cable, untwist no more than 0.5" of each pair of conductors; for Cat6 cable, untwist no more than 0.375" of each pair of conductors.
  • Make sure the riser cables are riser-rated or plenum-rated and that any cables that are laid in any part of the HVAC system (including ducts, return air spaces, etc.) are plenum-rated.
  • Avoid routing Unshielded Twisted Pair (UTP) cable near sources of Electro-Magnetic Interference (EMI), such as unshielded power lines, fluorescent lights, transformers, etc.  If a network cable must cross a power line, do so at a right angle.
  • Never lay UTP cable in the same conduit as power cables; doing so not only introduces huge amounts of EMI. Additionally, it's a violation of the National Electric Code (NEC) and very dangerous.
  • Communications equipment and the racks or enclosures that contain them need to be bonded to ground. Be careful when laying cable -- don't pull excessively (25 lbs limit) or tightly twist (1" Minimum Bend Radius (MBR)) the cable.
  • The maximum length of any Ethernet network segment over UTP is 328 feet.  This includes the patch cables between devices and Work Area Outlets, horizontal runs, and patch cables between ports on the patch panel terminating runs and ports on a hub or switch.  Generally, best practice is to plan for 16 feet of patch cable on each end (for a total of 32 feet) and limit horizontal runs to no more than 298 feet
  • The 802.3 Standard allows for up to three "populated segments" to exist in series (see the 5-4-3 simplification of the 802.3 10mbps model 1 Standard). 
  • Connections that originate in the main equipment room are called "Main Cross-Connects" and are usually abbreviated "MC" on network diagrams.  These correspond to the color "white" in the 606-A standard
  • Connections between wiring closets and work areas are called "Horizontal Cross-Connects" and are usually abbreviated "HC" on network diagrams.  These correspond to the color "blue" in the 606-A standard
  • Connections between equipment rooms other than the main equipment room and wiring closets are called "Intermediate Cross-Connects" and are usually abbreviated "IC" on network diagrams.  These correspond to the color "gray" in the 606-A standard
  • Label every cable in the network and maintain documentation of which cables go where.  There should be a label six inches from both ends of every cable that uniquely identifies the particular cable.  You should be able to refer to your documentation and know the exact path -- including which cables (and their labels) and the exact ports where those cables terminate -- without having to examine your equipment.
  • Network cabling is the core of your network; even with the best, most expensive equipment on the market, if the cable is of poor quality, the network will be unreliable, if it even works at all
Hopefully this information will help you better interact with your network and your network's support staff.  A reliable network is a key requisite to taking advantage of the cost-saving efficiencies of Google Apps and other cloud computing solutions.


-- Wes Dean, a Google Apps Certified Deployment Specialist and a Google Apps Trusted Tester, is Principal of KDA Web Technologies, a Google Apps Authorized Reseller. To learn how Wes and KDA Web Technologies can help you, go to http://www.kdaweb.com/.