Browse Prior Art Database

Original Publication Date: 1999-Apr-01
Included in the Prior Art Database: 2003-Jun-19
Document File: 5 page(s) / 79K

Publishing Venue



Main Idea for Disclosure RPS8-1999-0022

This text was extracted from a PDF file.
At least one non-text object (such as an image or picture) has been suppressed.
This is the abbreviated version, containing approximately 50% of the total text.

Page 1 of 5


Main Idea for Disclosure RPS8-1999-0022

Archived On: 05/05/99 03:20:55 AM

*** IBM Confidential ***

ABSTRACT: Normally, a properly written program will perform various types or work within it's domain and tell the user that something failed. The user will then have to interpet the fail based on documentation or quite often very cryptic error descriptions. In this invention the program creates an error file that describes which operation that failed plus any unique information to add more clarification to the failing operation. However, where it differs from current programs is that the error file can be used to recreate the failing work and the history of the events that led up to the fail in a unique method. The error file stores information in a way that can be transformed into a script file that is exactly the same format as program input files. That ability allows modification of the work and it's properties to improve time to fail.

If you can change the characteristics of a program fail to speed it up or better define it then you will minimize the time to resolving the problem. There are different advantages to using this technique depending on what the program is used for. For instance, a program discovers a bug within a particular computer subsystem. The design team for that subsystem is brought in. If the bug can easily be repeated then the team can probe various software/hardware elements to determine which caused it. Also, if they can modify the characteristis of the fail to improve definition then they may be able to solve bugs quicker and enable a faster time to market for that subsystem. Another example occurs when a customer uses a solution to perform work for their business. When the solution doesn't work properly the customer calls for service. If the program takes advantage of this approach the solution can be traced to a failing operation and reexecuted without recompiling the source. The customers software stack doesn't change and the service team has a mechanism to speed up the time for a fix.

The program uses a script file which defines the work that will be executed. This script contains global characteristics and one or more user defined threads consisting of one or more test objects. Test objects are the most primitive work element. When a test object fails it sends to the error handler specific information about the failing operation. This includes variables which define which test object in which thread that failed. For instance, a thread may have four test objects and the third failed. The user opens the debug view of the tool as shown in Figure 1. The errors were stored in an error file that the user can open via an error file by pressing the open button in Figure 2. After selecting one of the error files the view is updated with the corresponding error information as in Figure 3. The program correlates the variables stored within t...