Saturday, August 7, 2010
Ethernet
The Mike Meyers CompTIA Network+ All-in-One Exam Guide Fourth Edition is an excellent guide book you can use to study for the Network+ certification. Click on the Amazon 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 at New Horizons for the Network+ certification that you might find useful.
The Ethernet standard dominates networking technology today and has been around since Xerox developed it in 1973. The Ethernet is standardized as 802.3 and is based on the bus topology using twisted pair wiring.
Every version of Ethernet uses either unshielded twisted pair (UTP) or fiber. Every version of Ethernet uses a hub or a switch, hubs being very rare today.
Only 10 megabit and 100 megabit Ethernet use a hub.
1 gigabit and 10 gigabit Ethernet networks must use a switch.
Every version of Ethernet has a limit of 1024 nodes.
The maximum distance from the switch or hub to the node for UTP Ethernet is 100 meters.
Since 1990, Ethernet uses a hybrid star-bus topology and the hub acts like nothing more than a repeater. These ethernet hub networks use half-duplex and CSMA/CD for transmitting data. Every network breaks up data into pieces, called frames. This prevents a computer from hogging the shared cable. The data frames contain MAC addresses to identify the sending and receiving computer. CSMA/CD stands for carrier sense, multiple access/collision detection and determines which computer can access the wire and when. CSMA/CD is an 802.3 IEEE standard.
A basic Ethernet frame contains the following: the preamble, the MAC address of the receiving computer, the MAC address of the sending computer, the data length, the data itself, a pad, and a CRC (cyclic redundancy check, i.e., a frame check sequence).
Preamble: 64-bit series of alternating ones and zeros that end with 11. The preamble lets the receiving computer know a data frame is imminent.
MAC address: Every NIC has a unique MAC address. The hub sends out a copy of the frame to each computer on the wire. The computer with the matching MAC address recognizes the frame, while the other computers ignore it.
Length: A frame can carry 1500 bytes of data in a frame and the length field tells the receiving computer how many bytes of data are coming.
Data: The data part of the frame actually contains the data and also the IP address.
Pad: A pad is added to the frame if the data is less than 64 bytes.
CRC: The frame check sequence is like a check sum that enables the system to verify if all of the data was received. The sending computer computes the CRC by using a special mathematical formula and attaches the result to the frame. The receiving computer does the same thing, and if the CRC does not match, the receiving machine lets the sending machine know so that the data can be retransmitted.
The earliest version of Ethernet was 10Base5 and 10Base2, now obsolete.
10BaseT was created in 1990 and became the most popular networking standard in the world.
The number 10 refers to 10 Mbps and the word Base refers to baseband. The letter T refers to UTP cabling. 10BaseT requires the use of CAT3 or higher. CAT3 is a two pair wire, UTP (unshielded twisted pair) cable. One pair of wire sends data and one pair receives data. 10BaseT only requires two pair of wires, however, most people uss the four-pair wire to ensure against future technological advancements. The RJ-45 was also introduced at this time for 10BaseT. Each wire in the cable attaches to a pin in the RJ-45 connector. The real name for an RJ-45 is 8P8C, however, everyone calls it the RJ-45 connector. The TIA/EIA (Telecommunications Industry Association/Electronics Industries Alliance) defined the standard used called the TIA/EIA 568A and 568B.
10BaseT has a distance limitation of 100 meters. A 10BaseT hub can connect no more than 1024 computers.
10BaseFL was introduced a few years after 10BaseT. 10BaseFL is the fiber-optic version and transmits data using pulses of light. 10BaseFL signals can travel much further than 10BaseT electrical signals. The maximum length is up to two kilometers, and is immune to electrical interference (EMI). 10BaseFL uses multi-mode fiber-optic and uses either an SC or ST connector.
Media converters can be used to connect different kinds of cable, such as 10BaseT and 10BaseFL. Once the network expands in size, you need a way to connect efficiently. A bridge can be used to connect hubs or an uplink port or crossover cable.
With an uplink port, two hubs can be connected using a straight-thru cable. In order to connect two hubs, insert one end of the cable into the uplink port and the other end into the other regular port. You must daisy chain the ports, to connect multiple hubs.
Hubs can also connect using crossover cables. A crossover cable reverses the sending and receiving pairs on one end of the cable. This allows the two hubs to hear each other.
A bridge can be used to connect two Ethernet segments. The bridge acts like a repeater, but also monitors network traffic. Bridges are mainly used in wireless networks. Switches are mainly used in wired networks to connect segments.
An Ethernet switch looks like a hub but is intelligent. The switch uses the MAC addresses, creating point to point connections. Point-to-point is when two nodes have a dedicated connection. The switch creates an electronic table of the MAC addresses and acts like a telephone operator. Speed is enhanced and there is no longer a need for CSMA/CD. All devices connected to the switch will hear a broadcast sent from any one system, essentially a broadcast domain.
Spanning Tree Protocol was something that added a bit more intelligence to the switches to take care of the problem called bridge loop. The loop happened when there were too many switches which caused an endless bridge loop.
By the late 1990s, 100BaseTX became the most popular 100 megabit Ethernet standard. 100BaseTX is referred to as 100BaseT and uses CAT 5 cables. 100BaseT has a speed of 100 Mbps, a signal type of Baseband, 100 meters distance between the hub and node, CAT 5e or better UTP cables with RJ-45 connectors, and is a star-bus topology.
In a 100BaseT network, you must have a CAT 5 cable or better. Most network cables today are at least CAT 5, so this probably isn't an issue. You must have 100BaseT NICs and a 100BaseT hub or switch. 100BaseT NICs are called 10/100 NICs, and when the NIC first connects with the network it automatically determines the hub or switch speed. If the hub or switch only does 10BaseT, then the NIC does 10BaseT. If the hub or switch does 100BaseT, then the NIC does 100BaseT. 100BaseT NICs have totally replaced 10BaseT NICs.
Most Ethernet networks use UTP cabling, but there are many that are fiber-based because UTP cabling doesn't meet the needs of the network requirements. If the property is large, the 100 meter distance limitation doesn't work. UTPs unshielded twisted pair cabling containing a lack of electrical shielding does not work if there is a lot of electrical interference. Not to mention, it is easy to tap, and if you need a secure environment, 100BaseFX fiber-optic cable is better. 10BaseFX looks just like 10BaseFL, and has a maximum cable length of 2 kilometers, and is 100 Mbps. 100BaseFX uses ST or SC connectors.
By the end of the 1990s, the IEEE approved the 1000BaseT and the 1000BaseX standards. 1000BaseX is divided into other standards such as 1000BaseCX, 1000BaseSX, and 1000BaseLX. 1000BaseT uses four-pair UTP cabling and like 10BaseT and 100BaseT, has 100 meters maximum cable length. 1000BaseT is now the dominant Gigabit Ethernet standard.
1000BaseCX uses twinaxial cable with a length limit of only 25 meters. 1000BaseSX uses multimode fiber-optic cabling, and has a maximum cable length of 220-500 meters and rely on the SC type of connector. 1000BaseLX use single-mode laser cables up to 5 kilometers, and with a repeater, the distance can be increased up to 70 kilometers. 1000BaseLX connectors look like 1000BaseSX SC connectors.
10 Gigabit and 100 Gigabit Ethernet was soon developed and standardized by the IEEE members, to satisfy the increased desire for more bandwidth. In 2006, IEEE came up with a standard for 10 Gigabit Ethernet running UTP, 10GBaseT. Unfortunately, 10GBaseT, running on CAT6 cable has a maximum cable length of only 55 meters. CAT 6a cabling enables 10GBaseT to run at a distance of 100 meters.
Questions
1. Ethernet hubs take an incoming packet and repeat it out to the connected ports.
2. What begins an Ethernet frame? preamble
3. What type of bus does 10BaseT use? Star bus
4. What is the maximum distance for an ethernet segment? 100 meters
5. When used for Ethernet, unshielded twisted pair uses what type of connector? RJ-45
6. What is the maximum number of connected segments for 100BaseT? 1024
7. Which connectors are used by 10BaseFL cable? SC and ST
8. Which network devices can use the Spanning Tree Protocol (STP)? Switches
9. What device filters and forwards traffic based on MAC addresses? Bridge
10. What is the maximum distance between the hub/switch and node for 100BaseT Ethernet? 100 meters
11. What is the type of cable and connector that 100BaseFX uses? Multimode fiber with ST or SC connectors
12. How many pairs of wires do 10BaseT and 100BaseT user? 10BaseT and 100BaseT use two wire pairs
13. What is the standard IEEE 802.3ab? IEEE 802.3ab describes 1000BaseT which is known as Gigabit Ethernet and uses four pair UTP cabling to achieve gigabit performance.
14. What is the max distance for 1000Base LX with out repeaters? 5000 meters/5 kilometers. Some manufacturers increase the distance up to 70 kilometers with special repeaters.
15. What is the big advantage of using fiber-optic cabling? Fiber is not affected by electrical magnetic (EM) interference.
16. What is the standard connector for the 10-GbE fiber standard? There is no standard. It is left up to the manufacturers.
17. How many wire pairs does 1000 BaseT use? 1000BaseT uses four pairs of wires
18. What is the maximum cable length of 10GBaseT on CAT 6? 10GBaseT on CAT 6 cables are limited to 55 meters.
Tuesday, August 3, 2010
The OSI Model
Image via Wikipedia
Open Systems Interconnection (OSI) seven-layer model
The seven layers are:
- Layer 7 Application
- Layer 6 Presentation
- Layer 5 Session
- Layer 4 Transport
- Layer 3 Network
- Layer 2 Data Link
- Layer 1 Physical
Questions
1. Where does a hub send the data? Data comes into a hub through one wire and the data is sent out to all other connected wires to the other systems.
2. What uniquely defines every NIC? A NIC is a network interface card and what identifies the NIC is the MAC address, the media access control address.
3. What Windows utility is used to determine the MAC address? IPCONFIG /ALL from the command line.
4. A MAC address is known as what type of address? The physical address, a 48-bit value. It is a unique address assigned to each NIC. An example of a MAC address is 00-50-57-A4-05-0D.
5. A NIC sends data in chunks called what? Frames. Networks use frames to prevent one NIC from hogging the wire.
6. Which MAC address begins a frame? The MAC address of the receiving NIC begins a frame, followed by the sending NIC MAC address, and then followed by the data and CRC.
7. What does the CRC do? CRC stands for cyclic-redundancy check, a special bit of checking information that the receiving NIC uses to verify the data arrived correctly.
8. What layer of the OSI model controls the assembly and disassembly of data? The Transport Layer, the breaking up of data into chunks that are sent across a network and reassembled on the receiving end. Refer to the link above for a detailed explanation of the Transport layer.
9. What layer of the OSI model keeps track of a system's connections in order to send the right response to the right computer? The Session layer keeps track of a system's connections so that the right response will be sent to the right computer.
Image via Wikipedia
Related articles by Zemanta
- The Web Is Finally Evolving To A More Semantic Future (regulargeek.com)
- IP Transmission Technologies Hourglass of TCP/IP Protocols (slideshare.net)
Subscribe to:
Posts (Atom)