Fundamentals
of Networking
Chapter
4 Notes
Network
Topologies and Architectures
Topology – Map or physical layout of all the devices connected to the network. Two types of topology: 1.) Physical and 2.) Logical.
Access Methods - An access method is a set of rules for sharing the transmission medium. Types of access methods: 1.) Contention – first come, first serve, 2.) Polling – one device is responsible for polling (asking) the other devices to see if they want to transmit data, and 3.) Token Passing – computers take turns (must have the token) to transmit data.
Contention – Any device can send as long as the transmission line is not busy (first come, first serve). If two devices transmit at exactly the same time, a collision occurs. Both devices back off, wait a random number of clock ticks and retransmit again. CSMA/CD – Carrier Sense Multiple Access/Collision Detection – Tries to detect collisions before they occur. Although collisions do slow down network traffic, this is the most common network access method.
CSMA/CA – Carrier Sense Multiple Access/Collision Avoidance – reduces collisions by sending a warning broadcast.
Contention networks are called probabilistic because a computer’s chance of being permitted to transmit cannot be predicted. The more computers on the network, the more collisions will occur. All computers on the network are equal; meaning no priorities can be assigned. Contention is well suited for networks that have bursts of traffic.
Polling – One device (controller) is responsible for asking other devices if they want to transmit. Another words, devices cannot transmit until they are asked. This access method is not widely used because the act of polling adds extra network traffic.
Token Passing – A frame called a token circulates around the network. When a computer receives the token, it can then transmit data. Token passing is deterministic, since message is more predictable. When the computer is finished transmitting, it passes the token to the next computer in line. Token Ring network standards: 1.) IEEE 803.4 (Token Bus), 2.) IEEE 802.5 (Token Ring), and FDDI.
Token Passing can state priorities, and Token passing is more appropriate than contention under the following conditions:
Throughput - the speed at which data is passed through a network.
Comparing Contention and Token Passing – Token passing networks are more expensive than contention. Token Passing networks outperform Contention networks in high volume traffic, and Contention outperforms Token Passing with lighter volume traffic.
Network Topologies - Topologies can be described physically or logically. A physical topology is the actual appearance or layout of the network. The logical topology describes the flow of data through a network. There are many network topologies. Some of the most popular are:
Mesh
Star
Bus
Ring
Hybrid
Networks are considered to be Point-to-Point or Multi-point.
Metric Conversion - One Meter
equals 39.37 inches or 3.28 feet.
IEEE 802.3 (Ethernet) Media:
“In late 1972, Metcalfe and his Xerox PARC colleagues developed the first experimental Ethernet system to interconnect the Xerox Alto, a personal workstation with a graphical user interface. The experimental Ethernet was used to link Altos to one another, and to servers and laser printers. The signal clock for the experimental Ethernet interface was derived from the Alto's system clock, which resulted in a data transmission rate on the experimental Ethernet of 2.94 Mbps.
Metcalfe's first experimental network was called the Alto Aloha Network. In 1973 Metcalfe changed the name to "Ethernet," to make it clear that the system could support any computer-not just Altos-and to point out that his new network mechanisms had evolved well beyond the Aloha system. He chose to base the name on the word "ether" as a way of describing an essential feature of the system: the physical medium (i.e., a cable) carries bits to all stations, much the same way that the old "luminiferous ether" was once thought to propagate electromagnetic waves through space. Thus, Ethernet was born.”
Using Ethernet – (CSMA/CD) The network card listens to the network when it has a frame to send. If the card senses another card sending a frame at that moment, it waits and tries again. Ex CSMA/CD – Carrier Sense Multiple Access/ Collision Detection.
Collisions do happen and when they do, frames are re-transmitted. A collision occurs when two workstations try to transmit at the exact same time. The adapter card handles the re-transmissions of data, not the software. A collision takes one microsecond (one millionth of a second). Most of the time CSMA/CD Ethernet does not mean poor performance, faulty cables and cards do.
On the Ethernet network, data is broadcast to all machines at 10Mbps. Only those workstations that address matches network address in packet sends an acknowledgement. Suppliers of Ethernet hardware are Digital, 3 Com, Data General, Thomas Conrad.
Similar to True Ethernet is the Ethernet IEEE 802.3 LAN Standard. True Ethernet is based on network designed by Xerox in 1975. Unlike 802.3, true Ethernet uses one data link layer. Ethernet 802.3 LAN standard is not the original Ethernet and it uses 2 data link layers. These two differ by:
Ethernet Frames vs. 802.3 Frames:
True Ethernet Frame:
----------------------------------802.3 Frame:
Preamble: 8 bytes ----------------------------------Preamble: 7 bytes
Destination: 6 bytes --------------------------------SFD: 1 byte
Source: 6 bytes ------------------------------------Destination: 2 or 6 bytes
Type: 2 bytes -------------------------------------------Source: 2 or 6 bytes
Data: 46-1500 bytes -------------------------------Length:
2 bytes
CRC: 4 bytes ------------------------------------Data: 0-1500 bytes
-----------------------------------------------------PAD: filler
-----------------------------------------------------CRC: 4 bytes
True Ethernet Frame:
|
Preamble 8 Bytes |
Destination 6 Bytes |
Source 6 Bytes |
Type 2 Bytes |
Data 46-1000 Bytes |
CRC 4 Bytes |
The difference between the frames is the TYPE on the True Ethernet and the LENGTH on the 802.3.
802.3 (Standard) Ethernet Frame:
|
Preamble 7 Bytes |
SFD 1 Byte |
Destination 2 or 6 Bytes |
Source 2 or 6 Bytes |
Length 2 Bytes |
Data 0-1000 Bytes |
PAD |
CRC 4 Bytes |
Preamble: It’s used for synchronization and has 10101010 in 7 bytes (802.3), 10101011 in the 8 bytes (Ethernet).
SFD: (Start Frame Delimiter) - Preamble in Ethernet. Start Frame in 802.3. One is broken up and one is not.
Destination: It’s 6 bytes in Ethernet and 2-6 bytes in 802.3. It contains the address of the workstation that will receive the frame. If the first bit is a 0, then it will go to a single workstation. If it is a 1, then it will be broadcast to the entire network.
Source Address: Identifies the workstation that is sending the frame. It’s 6 bytes in Ethernet, and 2-6 in 802.3. The left most bit is always a 0.
Type: (It’s only in Ethernet, not in 802.3) Identifies the higher protocols that issue or wants to receive a frame. It enables higher protocols to share the network and not get into each other’s messages.
Length: (Only in 802.3) It tells the length of the data located in the frame.
Data Portion: It’s the message that’s used in both types of Ethernet. The only difference is that Ethernet’s is 46-1500 bytes and 802.3 it’s 0-1500 bytes.
PAD: (Only in 802.3) 0-1500 Added to Data Portion if length is less than 46 bytes for filler. Under True Ethernet, the packets that are too small are errors.
CRC: Cyclic Redundancy Checksum- It's the same in both. A formula is calculated by the workstation and compares it to the frame to see if the frame arrived intact or not.
NOTE: Since the frame sizes are different, this could make these two incompatible.
Token Ring:
Can be wired in a ring on a star configuration. A workstation sends a frame to a MSAU (multi station access unit, also MAU) which routes the frame to the next workstation. The frame moves in a ring from one workstation to another. The frame may be data being sent on a token. A token is a 3-byte message saying the LAN is idle. When a workstation wants to send, it must wait for the token. The frame goes from one workstation to another (adapter to
IBM has two versions, 4 or 16mbps. If it's faster than 16mbps, it's used for Ethernet. Some of the same vendors that make Ethernet also make Token Ring. Ex Thomas Conrad, 3 Com, Gateway, Proteon,
TI & IBM jointly developed TMS380, a chipset for Token Ring. IBM builds their own later.
One workstation on the ring is the "active monitor" who is responsible for issuing new tokens if necessary and generates error messages when necessary. The active monitor is chosen during initialization and can be any workstation on the network. If it goes down another workstation is on standby to take over.
IEEE 802.5:
Token Ring and LAN has three frame formats:
1. Token, 2. Data Frames, 3. Abort Sequence
Token:
3 bytes long lets workstation know when it is workstation's turn to send.
|
SD |
AC |
ED |
(Each field is one byte long)
First Byte (SD) - Start Delimiter
Second byte (AC) - Access Control
Third Byte (ED) - Field End Delimiter
Access Control Field
|
PPP |
T |
M |
RRR |
PPP - Priority Bit - (First three bits of the byte) Can range from 0 to 7, where 7 is the highest for a workstation.
T - Token Bit - is 1 for token and 0 for frame
M - Monitor Bit - Set to 1 by active monitor, set to 0 when workstation has it
RRR - Reservation Bit
If the active monitor sees a token with the monitor bit as 1, it knows that the token has been around the ring and not processed by a workstation. So the active monitor cancels it and sends out another token.
Token Data Frame-
/-------------SFS --------------/********************************/--------EFS-------/
|
DS 1 Byte |
AC 1 Byte |
FC 1 Byte |
DA 2 or 6 Bytes |
SA 2 or 6 Bytes |
Data 4500 Bytes Max |
FCS 4 Bytes |
ED 1 Byte |
FS 1 Byte |
MAC - Internal Message for network
LLC - computer to computer messages (data)
SFS: Start frame sequence is 3 fields
1. Start Delimiter - (same as in token) every frame or token starts with SD, it is 1 byte long.
2. Access Control - (same as for token)
|
PPP |
T |
M |
RRR |
PPP - Priority bit
T - Token bit
M - Monitor bit
RRR - Reservation bit
|
FF |
CCCCCC |
FF - Value 00 for MAC Frame and 01 for LLC frame
CCCCCC - Type of ring management
Destination Address (DA): 2-6 bytes long. If it has 2 bytes, then it identifies a group or individual address (same as CSMA/CD 802.3). With 6 bytes, it tells if address is logically or globally on the network. The rest of the bit tells the address of the workstation where it goes.
The Source Address (SA) – (sending device) same size as destination
The Data Portion - no maximum length (practical limits based on the workstation has control of the ring.
FCS - Frame Check Sequence - uses formula for error detecting (CRC).
End Frame Sequence - EFS has 2 fields These are:1. End Delimiter (ED) and 2. Frame Status (FS)
End Delimiter-has two fields, 1 bit each. A 1 is present if frame is part of multiple frame transmissions and 0 if last or only frame. Error Detection bit is set to 0 when a frame is sent originally. If a 1 is present, an error exists in the frame. The sending card then re-transmits the frame.
Frame Status (FS) - used to ensure data integrity.
|
A |
C |
RR |
A |
C |
RR |
Address recognized bit
0-when frame is originated
1-workstation sets it if address is recognized
3rd type of Token Ring Frame -
Abort Sequence - used to interrupt or terminate the current transmission.
|
SD |
ED |
SD - Start Delimiter
ED -End Delimiter
The Abort Sequence signal cancels the current frame or token transmission.
Using FDDI - (Fiber Distributed Data Interface) - fast speeds 100mbps and longer distances are the main differences. FDDI is faster than Ethernet or Token Ring and designed by ANSI.
The scheme used by FDDI is 4B/5B takes 4 bits into 5 bits for transmission so fewer signals are needed to send a byte of information. 5bits were chosen to ensure network-timing requirements are met. The 4B/5B at 100mbps actually causes 125 million signals per second. to occur (125 megaband). Hardware operates on 4 bits at one time vs. 1 bit at one time, therefore making it easier to have a high data rate.
Difference in Token Ring 802.5 and FDDI Token Ring -
In token ring (802.5), a new token is circulated only after a sending workstation gets back the frame it has sent. In FDDI, a new token is circulated immediately by the sending workstation after it finishes transmitting a frame. FDDI doesn’t use the Priority and Reservation subfields like token ring. FDDI classifies workstations into two classes:
1. Asynchronous -Workstations that are not rigid about the time periods that occur between network accesses (not synchronized)
2. Synchronous - Workstations that have timing requirements between accesses (synchronized).
FDDI Token
|
Preamble 8 Bytes |
SD 1 Byte |
FC 1 Byte |
ED 1 Byte |
FS 1 Byte |
Preamble-------------Start Delimiter--------Frame Control------- End Delimiter----- Frame Status Field
FDDI Frame
|
Preamble 8 bytes |
SD 1 byte |
FC 1 byte |
DA 2 or 6 bytes |
SA 2 or bytes |
Data 4500 max |
FCS 4 bytes |
ED 5 bytes |
FS 1 or 2 bytes |
Just like Token Ring frame carries user data or control network data.
1. Preamble - Used for synchronization and can be modified to workstation synchronization. (16 million=64 bits on Token Ring)
2. Start Delimiter (SD) - identifies the beginning of the frame.
3. Frame Control (FC) -
|
C |
L |
FF |
TTTT |
C - frame class synchronous or asynchronous
L - Frame address length (16 or 48 bits)
FF - Frame format, MAC- network data, LLC-user data.
TTTT - used if data is a MAC message.
A. DA – (Destination Address) - 16 or 48 bits long. Where data is to be sent.
B. Source Address - (SA) 16 or 48 bits long. Identifies the sending workstation
C. Data - (Information), contains data varies in length, but the entire frame can't exceed 4,500 Bytes.
D. FCS - Frame Check Sequence - 4 bytes CRC used for error checking
E. ED - (End Delimiter) - one nibble in a frame. One byte in a token. Both identify end of frame or token.
F. Frame Status - (FS) - any length contains the error detected bit, the address-recognized bit and the frame copied bit. Same on FDDI as Token Ring networks.
ARCNet - older architecture not found too often in business networks, but is found in older networks. ARCNet uses a star or bus topology normally with token passing, but does not require a MAU. ARCNET NICs have Dip switches that are used to configure the card. The maximum length between a node and a active hub is 610 meters, and the maximum length between a node and a passive hub is 30.5 meters. ARCNet can have a total of 255 stations per network segment.
Using NDIS and ODI -
NDIS - (Network Driver Interface Specification) and ODI- (Open Data Link Interface) - Two competing standards. Both determine how network operating system controls the adapter card (NIC). NDIS developed by 3COM and Microsoft. ODI developed by Novell and Apple.
NDIS- Warp server and NT server (Microsoft and IBM) the card manufactures can make their boards work with these operating systems by supplying NDIS drivers (system files) for the boards.
ODI- does same functions as NDIS but these are not compatible. Manufactures supply driver so cards will work. NIC manufactures supply both drivers so cards will work on either platform since they are not compatible.
(Novell vs. IBM) - MAJOR COMPETITION
PROTOCOLS -
Low Level Protocols - Ethernet, Token Ring, FDDI- Carry data from computer to computer but doesn’t handle file re-direction or file servers. Low level protocols don't ensure messages arrive in the same order as they are sent.
Mid Level Protocols - NETBIOS, IPX, SPX, TCP/IP (these fall into the Transport layer of the OSI model). These protocols do understand file re-direction and file servers.
For a glossary of Telecommunications and Networking terms visit http://www.visionael.com/services/glossary/index.html