Monday, August 16, 2010

Structured Cabling

Optical time-domain reflectometer Yokogawa AQ7...Image via Wikipedi


The Mike Meyers CompTIA Network+ All-in_One Exam Guide Fourth Edition is an excellent guide book you can use to study to obtain the Network+ certification.  Click on the adjacent Amazon link for a detailed synopsis of the book and an easy way to purchase it.

Below, are a few notes I have taken during my course of study for the Network+ certification that you might find useful.

TIA/EIA standards are the rules for structured cabling.  The organization BICSI (Building Industry Consulting Services, International), www.bicsi.org,  provides information about structured cabling and provides certifications for the cabling industry.







Cables run horizontally from the telecommunications room to the PC work areas.  This is called horizontal cabling.  A single piece of horizontal cabling is called a run.  The area on the opposite end of the cabling is called the work area (where the PC usually is).

Horizontal cabling is the cabling that runs from the telecommunications room to the work area.  The cabling is usually CAT 5e or better UTP cable, solid core or stranded core, very inexpensive and versatile.  Solid core has a single solid wire and is thicker.  Stranded core is a bundle of tiny wire strands and is more pliable.  Solid core will break if handled incorrectly, but is a better conductor, than stranded core.  Stranded core is much easier to handle than solid core.  Horizontal cabling should always be solid core because it will be installed in the ceiling and there is less risk of the cable being bent, or stepped on, or damaged in some other way.  Plenum cabling is used between walls because it is fire resistant.

The telecommunications room is also known as the intermediate distribution frame (IDF).  Because this is where all of the cabling comes together, the telecommunications room can develop into a messy monstrosity overtime if care is not taken to systematically organize and label the cables when a change is made.

You will need an equipment rack in a telecommunications room.  A patch panel is a box used to contain the cable connections with the short end of the cable going into the front and the longer end coming out of the back.  UTP cables connect to a 110 punchdown block in a structured cabling system using a punchdown tool. The metal-lined grooves for the individual wires and the punchdown tool forces the wire into the groove.  Place a label on the patch panel to identify each cable.  Patch panels come in varying types, for UTP, STP, or fiber ports.

Install the patch panel, and connect the ports to the switch using a patch cable.  Patch cables are UTP stranded cables, and easier to handle than solid core.

In the work area is the wall outlet that consists of one or two female jacks.  This jack is the termination point for the network cable.  The PC is connected to the wall outlet with a stranded patch cable.  It is a good idea to label the wall outlet so that the wall outlet shows its position on the patch panel.

A demarc is the term used to describe the location where connections come into the building from the outside.  This location marks the dividing line of responsibility for network functionality.  The internal side of the demarc is your responsibility.  The other side of the demarc is the ISPs responsibility, or the company that provides the upstream service to you.

In your home, the DSL or cable modem is your demarc.  This is the line between your side and the service provider.  In the office, the demarc is more difficult to determine because of the increased layout complexity.  Most NIDs (Network Interface Device) today are known as smart jacks, because they enable the ISP to determine if the customer has disconnected from the NID.   Smart jacks also do remote loopback-good for loopback testing when the other end of the connection is some distance away.

A cable drop is the location where the cable comes out of the wall.  Before you map your runs, you should obtain a floor plan. There are external raceway products that adhere to the walls.  Raceways are good for older buildings where you don't want to open up the walls.

The room must be placed in a spot that won't require cable runs of more than 90 meters.  Plan the types of cable and runs needed.  Calculate your cable runs and add 25%.  Measure the vertical drop from the ceiling to the wall jack and ad 18".  Minimize sharp bends.
Make sure your telecommunications room has its own dedicated power circuit.
Have a wiring map.
Hire a certified company to do your wiring tests.
Make sure the room has low humidity and is well air-conditioned.
Remember heat is the enemy of computer equipment.  Keep the telecommunications room cool!
Make sure the telecommunications room is secure.

Local codes and TIA/EIA and the National Electrical Code (NEC) contain the rules of how to pull cable through a ceiling.  Each cable must be tested to ensure every connection works as desired and can handle the speed of your network.  TIA/EIA has certain standards for testing cable runs.  Testing should be done by a professional cable installer.

There are a few inexpensive tools a tech can use to test connections.  A multimeter tests continuity and is done by placing the probes on either end of the cable.  Set the multimeter to its continuity setting or to Ohms.  Electrical impedance is measured in ohms.  With Ohms, if you have a connection, you get zero Ohms, and if you don't have a connection, you get infinite Ohms.

A TDR, Time Domain Reflectometer, tests continuity and includes the additional capability to determine the length of the cable.  Some have a loop back tester to work with installed cables.  Good testers can test crosstalk and attenuation.

Twisted pair (TP) cable connections have four pair of wires.  Each pair is twisted to prevent crosstalk or EMI (Electrical Magnetic Interference).  Highly rated wires have more twists.  Crosstalk is when a signal is sent down one of the wire pairs, and the other pairs pick up some of the signal.   Near-end crosstalk (NEXT) is when an electronic detector device, picks up on the signal from the other pair on the end closest to the device.  If you repeat the test, sending the signal down the pair of wires on the far end of the connection, picking up a signal from the other wire, you have far-end crosstalk (FEXT).
Attenuation is when the signal progresses down a piece of wire and becomes steadily weaker.  This happens as the cable run gets longer. The maximum run for TP is 100 meters.
The process of verifying that every cable run meets the TIA/EIA standards requires testing tools, called cable certifiers.  Certifiers can do the testing and generate a report to show the customer.  Fluke networks, www.fluke.com, has electronic test tools and software.

Most fiber optic runs use only two cables, so they don't experience crosstalk/EMI.  However, they do break and require an OTDR (optical time domain reflector).  OTDRs are used to verify impedance and cable termination.  OTDRs tell you exactly how far down the cable to look for the break.  Fiber certifiers tests fiber optic runs to ensure a fiber will carry the designated signal speed.

The three big issues with fiber are attenuation, light leakage, and modal distortion.  Light diffuses over distance, causing attenuation, when the light signal spreads and the signal weakens.  Light leakage occurs when if the cable is bent too much, exceeding the maximum bend radius.  Multimode fiber has the problem of model distortion.  The light is sent out in different modes, some light shoots straight, and some bounce back and forth at a sharp angle down the fiber.

A NIC provides the physical connection between a workstation and the network infrastructure.  All UTP Ethernet NICs use the RJ-45 connector, and the cable runs from the NIC to the hub or switch.

There are many different types of fiber-optic NICs.  Check the documentation to determine what you need and opt for the multispeed cards.  You need to have the right connectors and protocol.  The MAC address is coded into all NICs.

Most PCs come with built-in NICs.  Some laptops have NICs located on the motherboard, which would require a motherboard replacement.  Most PCs come with two types of expansion slots, the Peripheral Component Interconnect (PCI) type and the PCI Express (PCIe).  PCIe is the newer type.  You can get NICs with a USB or PC Card connection.  USB is convenient, but has a maximum speed of 480 Mbps, and the PCMCIA/PC card is only good for laptops.

After you physically install the NIC, installing the NIC driver into the OS is the next step.  Verify the computer has the NIC and is in good working order.  On Windows, check Device Manager.  Ubuntu Linux has the Network applet, and Macs have the Network utility in System Preferences.

Bonding is when you use multiple NICs for a single machine.  This is also called link aggregation.  Bonding or link aggregation will at least double the speed between a machine and switch.

UTP NICs have light-emitting diodes (LEDs) that give info about the NIC's link.  These lights give you pointers about what is happening and is one of the things to check when you can't get a connection to the network.  Be sure to check the documentation for the meaning of the lights.  Switches and hubs also have link lights.  A very important light is the activity light.  The activity light indicates the network is functioning.  So, even if the link lights look good, if you are doing work on the computer and the activity light is not flckering, you know something is wrong.

Fiber-optic NICs are much more difficult to diagnose.  Many fiber-optic NICs don't have lights.  An optical tester enables you to check the connectors on a fiber-optic NIC.

If you receive the message, "No server is found" error, this is might mean you have a physical problem with the network.  First check for software errors.  However, a bad NIC can generate this message, so you should use the utility provided with your OS to verify the NIC works.  Some NICs have diagnostic software you can run.

If you believe the problem is hardware, first check the link lights on the NIC and switch.  If they are not lit, then you know the cable might be disconnected.  You can always check the network icon in your OS to determine if you have a connection, and go from there.

Make sure other people can access the problem server.  Inspect the cable running from the PC to the outlet.  If that doesn't work, try another outlet.  Most network disconnect problems occur in the work area, but, if you have tested and determined the problem is not in the work area, it could be time to test the horizontal cabling run.  A midrange tester with TDR like the Fluke Microscanner can be used to determine whether and where the cable is disconnected.

If your UPS comes on too often, it might be a good idea to install a voltage event recorder.  A voltage event recorder plugs into your power outlet and tracks the voltage over time.  Because heat is an issue with electronic equipment, your telecommunications room should have a temperature monitor.  Tracing cables will eventually become a need and a toner is a good device to have for that.  The toner is really a tone generator and a probe.  The tone generator connects to the cable using tiny hooks and sends an electrical signal along the wire at a certain frequency.  The tone probe emits a sound when it is placed on a cable connected to the tone generator, indicating the right cable.

A good cable tester and a good toner are the most important tools used by a network technician whose job it is to support a network.


Questions

1.  Which cable should never be used in a structured cabling install?  Coax

2.  Which type of network is a dedicated line allocated for the transmission of data between two network nodes?  Circuit switching

3.  What enables you to use multiple NICs together in a computer to achieve a much faster network speed?  Bonding

4.  The CAT 5e rating defines how many pairs of wires in the cable? 4

5.  What organizes and protects the horizontal cabling in the telecommunications room?  Patch panel

6.  What are patch cables used for?  They are used to connect PCs to outlet boxes, and, to connect the patch panel to the hub.

7.  Which network technologies use UTP cabling in a star topology?  10BaseT and 100BaseT

8.  Tina needs to increase network throughput on a 10BaseT network that consists of 1 hub and 20 users.  What is the most inexpensive hardware solution?  Replace the hub with a switch.

9.  What two devices together enable you to pick a single cable out of a stack of cables?  A tone generator and a tone probe work.

10.  Rack-mounted equipment has a height measured in what units?  U

11.  What is the most popular type of cabling currently used?  Fiber

12.  What device helps to prevent network collision by not allowing data to be broadcast?  A switch

13.  What cable type with a single copper wire is used for sending and receiving data?  Coaxial

14. What is the lowest grade of cable to run 1000BaseT?  UTP CAT5e

15.  What is the maximum distance to run UTP cable?  100 meters.  Thin-net coaxial is 185 meters.  Multimode fiber is 400 meters.


Enhanced by Zemanta