Chapter 1 – Token Rings

Download Report

Transcript Chapter 1 – Token Rings

Chapter 1 – Token Rings
Monday, January 22, 2001
Token Rings
• Originally proposed by Olaf Soderblum in
1969
• IBM purchased the rights and released its
product in 1984
• IEEE 802.5 standard finalized in 1985
• Operates at 4-16 Mbps
Features of Token Ring Network
• Highly reliable network
• More complex than Ethernet
• 802.3 standard is physically a star but is logically
a ring
• Uses a “free” token that is passed around the ring
in one direction
• If a node receives the “free” token, it can change
the token from “free” to “busy” attach data and
send it onto the ring
• Each station receives the data and sends the data
and token to the next node
Token Ring – Cont.
• The receiving station reads the data and sends it to
the upper level protocols
• The token is marked to indicate that the data has
been received (2 bits are switched)
• The token with data is then retransmitted and
finally gets back to the sender
• This serves as an acknowledgement that the data
was received
• This process is a “collision avoidance” system
Fault Management Techniques
• No circulating token
• Persistent busy token
• To address these, one station is designated
as the active token monitor
• See handout for details
Advantages of Token Ring
• Operates under heavy load with
significantly superior performance to
Ethernet
• Known upper bound to the amount of time
any station must wait before transmitting
• Built-in diagnostic and recovery
mechanisms (beaconing, autoreconfiguration)
Disadvantages of Token Ring
• Token ring cards and equipment are more
expensive than Ethernet
• Can be very difficult to troubleshoot
• Requires considerable expertise
Operation of Token Ring – See
Handout
• Eight levels of priority
• Priority field
• Reservation fields
Token Ring (802.5) Frame
Format
•
•
•
•
•
802.5 Header
802.2 Header (Logical Link Control)
Data
802.5 Trailer
See p. 16 for specific fields
802.5 Header
• Start Delimiter – 1 byte – provides
synchronization for the receiver
• Question –Why is only one byte used here instead
of the 8 bytes used by Ethernet II?
• Access Control – one byte –
– 3 bits for priority of token
– 3 bits for reservation level. These set the priority of the
token once the token is released
– 1 bit to indicate the ring monitor station has been
passed. What does ring monitor do if it sees this bit set?
– 1 bit to indicate whether the frame that follows is a
token or a frame
Frame Control
• One byte
• Indicates whether the frame that follows is a
management frame or an LLC frame (2 bits)
• Indicates the type of management frame
Destination Address
•
•
•
•
•
6 bytes long
Unicast address
Universal broadcast address (all 1s)
Token ring broadcast (C0-00-FF-FF-FF-FF)
Multicast address
Source Address
• 6 bytes
• Indicates the sending node’s address
Payload
• No minimum frame size
• Maximum size depends upon bit rate and
token holding time
Frame Check Sequence
• 32-bit CRC
• Access and Frame Control bits not included in this
process, so these bits can be changed without
recalculation of CRC bits
• These bits are checked as it passes each node in
the ring
• If the FCS fails at any node, the Error bit is set and
the receiving node doesn’t copy the frame
FDDI
•
•
•
•
•
•
•
Network developed by ANSI
Uses fiber optics
Token ring that operates at 100Mbps
Designed to operate over long distances
Usually acts as a high-speed backbone
Allows for several frames on ring
Station can release token as soon as it is through
transmitting frames
• Uses two counter-rotating rings
Homework
• Begin reading Chapter 2
• Be ready for possible quiz – Chapter 1