Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Dynamically Detecting Duplicate Node Names

IP.com Disclosure Number: IPCOM000113553D
Original Publication Date: 1994-Sep-01
Included in the Prior Art Database: 2005-Mar-27
Document File: 2 page(s) / 126K

Publishing Venue

IBM

Related People

Bertin, O: AUTHOR [+8]

Abstract

The joining of the two CP spanning trees as described in (*) begins by the lower-named root sending a Combine Request to the higher-named node. If the higher-named root is not in the process of combining with another CP spanning tree, the higher-named root sends a Combine Grant message to the lower-named root. At this point in the protocol the two roots would exchange topology information with each other, but this paper introduces an additional function that each root performs prior to sending the topology information--this function detects the presence of duplicate node names.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 44% of the total text.

Dynamically Detecting Duplicate Node Names

      The joining of the two CP spanning trees as described in (*)
begins by the lower-named root sending a Combine Request to the
higher-named node.  If the higher-named root is not in the process of
combining with another CP spanning tree, the higher-named root sends
a Combine Grant message to the lower-named root.  At this point in
the protocol the two roots would exchange topology information with
each other, but this paper introduces an additional function that
each root performs prior to sending the topology information--this
function detects the presence of duplicate node names.

      Prior to the two roots exchanging their network topology with
each other, each root constructs a Node Name List message to send to
the other root.  Each root puts all nodes which are in its CP
spanning tree in the Node Name List message and then sends the
message to the other root.  Upon receipt of the Node Name List
message, each root compares all received node names with the node
names it sent to the other root.  During this comparison, each root
determines whether the two CP spanning trees that are joining have
any duplicate node names.  If either or both roots find one or more
duplicate node names, the CP spanning tree join process is aborted by
sending a Combine Abort message with a sense data value of "duplicate
node name."  Each node whose name is a duplicate is also included in
the Combine Abort message.  The two CP spanning trees do not join to
form a single CP spanning tree in this case, unless network
management decides to rename the nodes with duplicate node names.

      Network management within the node also receives the sense data
and node names.  Network management then notifies the network
administrator of the duplicated node names.  By notifying the network
administrator of the duplicate node names, the network administrator
has the ability to determine what action to take, based upon the
current network and the customer's requirements.  The actions may
include any of the following:
  o  Take the node whose name was duplicated out of the network,
     rename the node, and then re-enter the node and its resources
     into the network.

          For example, the network administrator would take this
    action when the node has a minimal number of links attached to it
    but would not take this action if the node is a main hub of
    traffic.  Some items taken into account when deciding whether to
    take this action are as follows:  the amount of overhead involved
    in re-entering the node and its resources into the network; the
    need for the networks to become one (e.g., accessing the
    customer's requirements); and the amount of time allotted to
    joining the two networks (if the customer's business needs
    require that they become one).
  o  Wait for the other network administrator to rename its node with...