International Business Machines Corporation

John Majikes

Lap Thiet Huynh

Dinakaran B Joseph

Roberto Delima

Sivaram Gottimukkala

Robert E Moore

Lee M Rafalow

Jason P Hawrysz

Roger A Bess

Dakshi Agrawal

Douglas Morgan Freimuth

Roy Frank Brabson

David P Olshefski

Gary Owen McAfee

Janet C Wolf

Dinakaran D Joseph

A computer-implemented method of selecting among a plurality of endpoint policy alternatives to apply to a message conveyed over a data communications network is provided. The method can include assigning a score to each of the plurality of endpoint policy alternatives, wherein an assigned score is based upon policy assertions of the endpoint policy alternative to which the score is assigned. The method can further include selecting, according to a predetermined selection criterion, one of the plurality of endpoint policy alternatives based upon the assigned scores.

System and methods for efficiently classifying and selecting among security policy alternatives for outbound network communications

An invention is disclosed for improved computer network efficiency in use of remote procedure call (RPC) client/server applications functioning within dynamic virtual networks. Specifically, a method and system are disclosed for use of an RPC mapping program to provide a client with the network communications protocol address(es) having the highest probability of successfully establishing a connection with a server, in order to maximize the probability that the communications interface between client and server will be created in the minimum number of calling attempts.

Method and system for improved computer network efficiency in use of remote procedure call applications

A method, apparatus, and computer implemented instructions for processing a request in a data processing system. The request is received. In response to a first hash value being present within the request, the first hash value is compared to a second hash value that was computed locally, wherein the second hash value represents a current policy configuration for assigning a quality of service. In response to a match between the first hash value and the second hash value, other information in the request is used to establish a quality of service for packets associated with the request.

Policy-based packet classification

Techniques are disclosed that enable an application to prioritize its traffic in an application-specific manner, such that traffic which is most important to this application can be prioritized over traffic that is less important to this application, where this prioritization information is reflected in differential treatment of the traffic within the network. An application conveys relative priority information for its traffic to a Web proxy or edge server, and this device then consults stored policy information to determine how the traffic should be prioritized in the network. The Web proxy also ensures that usage of network resources is balanced among multiple applications. Cookies are used for transmitting priority information between the edge server and application in an efficient manner.

Application prioritization in a stateless protocol

Method and apparatus for policy-based packet classification using hashing algorithm

A fast policy classification method and system. A policy differentiator can be provided which can include a policy repository configured to store hierarchical policy classification trees. The policy differentiator further can include a hashing module configured to produce integer values for string based selector attributes. In one aspect of the invention, the hashing module can include a longest string hashing function. A policy matching engine can be provided which is configured to locate policy rules in the hierarchical policy classification trees according to the integer values produced for the string based selector attributes. Finally, a policy matching engine interface can be provided through which policy queries and resulting policy decisions of the policy matching engine can be communicated between the policy matching engine and the policy enforcement entity.

Fast policy classification for strings

Growing community of inventors

Apex, NC, United States of America

John Majikes