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

Efficient Scheme of Distributed Name Service

IP.com Disclosure Number: IPCOM000109064D
Original Publication Date: 1992-Jul-01
Included in the Prior Art Database: 2005-Mar-23
Document File: 4 page(s) / 81K

Publishing Venue

IBM

Related People

Nakajima, A: AUTHOR

Abstract

Disclosed is a scheme of distributing name data, which is a pair of name and address of a resource, to nodes in distributed name service and its name service protocol. In this environment, all nodes work as name servers, and each node has the same amount of load. This scheme uses minimum number of messages required in distributed name service when frequency of registration and that of query are not the same.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 55% of the total text.

Efficient Scheme of Distributed Name Service

       Disclosed is a scheme of distributing name data, which is
a pair of name and address of a resource, to nodes in distributed
name service and its name service protocol. In this environment, all
nodes work as name servers, and each node has the same amount of
load.  This scheme uses minimum number of messages required in
distributed name service when frequency of registration and that of
query are not the same.

      A name service algorithm requires associating with a node i two
subsets of nodes P(i) and Q(i), such that
      P(i)   Q(j) * /   .   (1 & i, j & n)
where n is the total number of nodes 1.  In the name service
algorithm, node i sends the address of a resource to nodes denoted by
P(i) for registration when a process in node i creates the resource
(Fig. 1).  The nodes in P(i) store the received name data.  There is
another process in node j, which wants to know the address of the
resource.  Then node j asks the nodes denoted by Q(j) for this
information.  Because the intersection of P(i) and Q(j) is not empty,
nodes in P(i)   Q(j) can send the address of resource to the process
in node j.  This is a simple abstraction of name service in
distributed systems, but it can support any types of name service by
constructing appropriate P(i) and Q(i).  The number of messages
required when node i registers name data is  P(i), and the number of
messages required when node i queries name data...