Method and apparatus for testing software

Abstract

Techniques are provided for testing new software slated to be deployed on a target machine population with a number of target machines having a number of existing software programs. Method steps can include obtaining a list of depending software programs on selected ones of the target machines. The list can include those of the existing software programs which depend on the new software and / or those of tile existing software programs on which the new software depends. The steps can further include identifying relevant test cases based on the list of depending software programs, instantiating a plurality of test client platforms characteristic of the target machine population, and testing the relevant test cases on the plurality of test client platforms to identify failures.

Description

FIELD OF THE INVENTION [0001] The present invention generally relates to information technology and data processing, and, more particularly, to a method and apparatus for testing software. BACKGROUND OF THE INVENTION [0002] Deploying new software to a population of client platforms is a costly process for both software vendors and their enterprise customers, due to the common occurrence of failures. Recovery from such failures usually entails platform users calling a support center multiple times; when problems nevertheless remain un-resolved, technical support personnel may be dispatched to the user work place. One reason for the prevalence of deployment failures is neglecting to test the new software against the failed client platform configuration. The inability to test new software against pertinent possible client platform configurations may be a problem for the software vendor, but is likely far less problematic in the case of an enterprise. An enterprise usually knows the con...

AI Technical Summary

Benefits of technology

[0009] One or more embodiments of the invention may provide one or more beneficial technical effects, such as, for example, identification of an appropriate (e.g., minimal) set of client platform configurations for use in testing and a set of test cases; because the selection can potentially be from among the entire population of client platform configurations targeted for deployment, the chance of a deployment failure caused by an untested client platform configuration can be reduced. Furthermore, to reduce testing time, new software can be tested against a set of test client platforms that emulate client platform configurations identified in the appropriate set of configurations. These test client platforms can be built, e.g., on physical computers or created as virtual platforms on a powerful server node.

Problems solved by technology

Deploying new software to a population of client platforms is a costly process for both software vendors and their enterprise customers, due to the common occurrence of failures.

One reason for the prevalence of deployment failures is neglecting to test the new software against the failed client platform configuration.

The inability to test new software against pertinent possible client platform configurations may be a problem for the software vendor, but is likely far less problematic in the case of an enterprise.

A deployment failure can occur, for example, when the newly deployed software failed to install properly or to perform as expected, or caused problems to the existing software on the target client software platform.

However, the Laviolette et al. reference does not teach how one might efficiently determine an appropriate set of client platforms and test cases to test for deployment.

Images