Brought to you by EarthWeb
IT Library Logo

Click Here!
Click Here!


Search the site:
 
EXPERT SEARCH -----
Programming Languages
Databases
Security
Web Services
Network Services
Middleware
Components
Operating Systems
User Interfaces
Groupware & Collaboration
Content Management
Productivity Applications
Hardware
Fun & Games

EarthWeb Direct EarthWeb Direct Fatbrain Auctions Support Source Answers

EarthWeb sites
Crossnodes
Datamation
Developer.com
DICE
EarthWeb.com
EarthWeb Direct
ERP Hub
Gamelan
GoCertify.com
HTMLGoodies
Intranet Journal
IT Knowledge
IT Library
JavaGoodies
JARS
JavaScripts.com
open source IT
RoadCoders
Y2K Info

Previous Table of Contents Next


Addition of Station

New stations are admitted to the logical ring through a response window—based controlled contention process. According to this scheme, a station waits a predetermined interval after the transmission of a media access control (e.g., a solicit-successor) frame and listens for a response. If the station detects a response at any time during this response window, it continues to listen beyond the duration of the response window. Duration of response windows are identified by two solicit-successor frames; solicit-successor 1 indicates one response window, and solicit-successor 2 indicates two response windows. The solicit-successor frame indicates the range of station addresses between the frame source and destination address. Any station or stations whose addresses fall within this range may indicate desire to enter the logical ring. Three events can happen during the response window:

  No response. No station expressed an interest in joining the ring. The token holder then forwards the token to the successor.
  One response. Only one station issued a set-successor response frame. The token-holding station then designates the responding node as the current successor and passes the token to the new successor. The responder then sets its linkages (i.e., predecessor and successor) accordingly and proceeds as usual.
  Multiple responses. The token holder detects a contention, indicating that more than one station intends to join the ring. The conflict is resolved by an address-based contention scheme using an algorithm wherein the sending node transmits a resolve-contention frame and waits for four response-window intervals. Each contender can respond during one of these windows, with responses based on the first 2 bits of its address. A contending station must refrain from transmission until the arrival of its window. If the token holder receives a valid set-successor frame, it passes the token to the appropriate station. If there is still a multiple response, the contention process is repeated. Only those stations that responded during the first window, however, are allowed to contend on the basis of the second pair of bits in their address. The algorithm continues until a valid set-successor frame is received. If there is no response or the maximum number of retries is reached, the sending node passes the token to the previously designated successor.

Deletion of Station

A station N must wait for a token before sending a set-successor frame to its predecessor N - 1 declaring its intention to drop out of the ring. Departing station N then passes the token to the successor N + 1 as usual. Receiving station N + 1 then automatically sets N - 1 (the predecessor address) as the source address of the token frame, indicating the token’s path was from N - 1 to N + 1. This splices the departing station out of the logical ring.

Fault Management

The fault management routines of the IEEE 802.4 token passing standard are more complex than those of the CSMA/CD 802.3 standard. The 802.4 fault management must include contingencies illustrated in Exhibit 1-1-23.


Exhibit 1-1-23.  IEEE 802.4 Fault Management Contingencies

Multiple Token

The multiple token condition occurs when more than one station holds a token, for example, in the case of two stations having the same address. During the token holding mode, if station N detects a frame indicating that another station also holds a token, node N drops the token and enters into the listening mode.

Unacceptable Token

A sending station passes the token to its successor, then spends one time slot sampling the channel. A valid frame activity on the channel during the interval means that the would-be successor is alive, while invalid or no activity means that the successor is dead or the token was unacceptable. The station then reissues another token to the same successor and waits one time slot for a valid response. If no valid response is detected after the second try, the station assumes that the successor is dead and proceeds to find a new successor.

Who Follows

To find a new successor, a token holder issues a who-follows frame, asking the identity of the station that follows the failed station. In this case, the token holder should receive a set-successor frame from the second station down the ring. On receipt of a set-successor frame, the sending node forwards the token and updates the address of its successor.

If the sending station does not receive a valid (set-successor) response to its who-follows message, it retries. After two failures, the station issues a solicit-successor frame containing a full address range, thus inviting any station to become a successor. If a successor is found (i.e., if there is a response), the ring resumes usual operation.

Should there be no response to the solicit-successor frame, the token holder assumes that a major fault has occurred in the network or that all nodes have dropped out of the ring or its own receiver has failed. The station then sends any remaining data and attempts to pass the token again. It then refrains from transmission and enters into listening mode.

Ring Initialization

Initialization of the ring is triggered when a station’s inactive timer expires (i.e., the token has been lost). This can be caused by startup of the station, crash of a token holder, and other network events. The initialization algorithm is triggered when a station issues a claim-token frame. This frame’s information field length is a multiple of the system slot time and is calculated on the basis of the first 2 bits of the issuer’s address. After transmission, a claimant listens for a valid response for one slot time from its own transmission and those of other stations that choose the same frame length. The station then samples the medium. In the event of a valid response, the station either drops its claim or tries again, using the second pair of its address bits. If all bits have been used and silence is still sensed, the station has succeeded in the initialization contest and becomes the token holder. The logical ring can now be created by using the response window process.


Previous Table of Contents Next

footer nav
Use of this site is subject certain Terms & Conditions.
Copyright (c) 1996-1999 EarthWeb, Inc.. All rights reserved. Reproduction in whole or in part in any form or medium without express written permission of EarthWeb is prohibited. Please read our privacy policy for details.