Browse Prior Art Database

Methodology to Identify Redundant Requests from Multiple Applications

IP.com Disclosure Number: IPCOM000107518D
Original Publication Date: 1992-Mar-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 2 page(s) / 107K

Publishing Venue

IBM

Related People

Johnson, WJ: AUTHOR [+3]

Abstract

In a cooperative processing environment, remote applications may propagate application requests to other remote workstations. Often network connections between remote applications are not known by the cooperating applications or services. This can result in an application receiving the same request from multiple applications. This article describes a method which allows a receiving application to identify if a particular request has been previously processed by the application. The article also describes a methodology whereby a shortest path can be taken to the originator of a request rather than taking the path which was used to propagate to the service which satisfied the request.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 51% of the total text.

Methodology to Identify Redundant Requests from Multiple Applications

       In a cooperative processing environment, remote
applications may propagate application requests to other remote
workstations. Often network connections between remote applications
are not known by the cooperating applications or services. This can
result in an application receiving the same request from multiple
applications. This article describes a method which allows a
receiving application to identify if a particular request has been
previously processed by the application.  The article also describes
a methodology whereby a shortest path can be taken to the originator
of a request rather than taking the path which was used to propagate
to the service which satisfied the request.

      A service is provided by the receiving system for maintaining
records for purposes of determining an identical request being
propagated by more than one system. Each receiving system maintains a
local table with entries which define pending requests and the
originators of those requests. The record is defined as:

                            (Image Omitted)

      The PROCESS ID contains unique information for the pending
request. The PROCESS ID contains the network address for the
originator of a request, an application identifier and a unique
requester identifier. The unique requester identifier uniquely
identifies the requesting component in the requesting application
identified by the application identifier. Thus, the tuple of
originating network address, an application identifier and a unique
requester identifier for the application uniquely represent the
originator of a request. A unique current date/time stamp is derived
by adding the originator's machine system time to a configured
timeout value. This timestamp is placed into the TIMESTAMP field. The
TIMESTAMP field represents the date and time of when the request
search expires (i.e., determined as could not satisfy because of a
timeout). The PROCESS ID field in conjunction with the TIMESTAMP
field uniquely identifies any request in a network.  The PROCESS LIST
field contains all nodes which have been propagated to, in search for
a node which can satisfy a request. The first node in PROCESS LIST is
always the originating node of a request. The re cord is maintained
in a local table to the receiving service machine.  All fields i...