Browse Prior Art Database

Optimization of Network Resource Filtering

IP.com Disclosure Number: IPCOM000107384D
Original Publication Date: 1992-Feb-01
Included in the Prior Art Database: 2005-Mar-21
Document File: 1 page(s) / 30K

Publishing Venue

IBM

Related People

Lillie, BT: AUTHOR [+3]

Abstract

Disclosed is a method for optimizing network resource filtering by reducing the amount of network traffic. For a random set of resources, instead of processing each request singly, all of the resource check requests for a particular server are buffered and sent to the server which owns the resources.

This text was extracted from an ASCII text file.
This is the abbreviated version, containing approximately 100% of the total text.

Optimization of Network Resource Filtering

       Disclosed is a method for optimizing network resource
filtering by reducing the amount of network traffic.  For a random
set of resources, instead of processing each request singly, all of
the resource check requests for a particular server are buffered and
sent to the server which owns the resources.

      Checking a set of resources to see if a user has access
requires that each resource be checked individually through an API
call.  If the resource is on a server, then the client must remote
the API to the server.

      To significantly reduce the number of network accesses made,
the list of resources to be checked is first scanned to build a
buffer of all resources for one server.  This buffer is then passed
to the server via a named pipe.  A process on the server will then
take the buffer, perform the API call to check access to the
resource, and put the result back in the buffer.  The buffer will
then be returned via named pipe to the client, who will unpack the
buffer to process the results.  Each server with resources will be
handled the same way.

      The number of network accesses has been reduced significantly
from one call per resource, to one call per server with a resource.
Also, calling an API locally is more efficient than remoting the API,
so additional improvement is realized here.