Dismiss
InnovationQ will be updated on Sunday, Oct. 22, from 10am ET - noon. You may experience brief service interruptions during that time.
Browse Prior Art Database

Error messages with unique identifiers

IP.com Disclosure Number: IPCOM000014412D
Original Publication Date: 2001-Jun-16
Included in the Prior Art Database: 2003-Jun-19
Document File: 2 page(s) / 45K

Publishing Venue

IBM

Abstract

Today's complex software products require advanced Problem Determination facilities such as trace and error/informational logging. These facilities are used to keep track of any important errors, status or informational messages which may be pertinent to the workings of the software product. These facilities are not usually well connected to any user interfaces and problem determination for visible errors often entails examining error logs or traces to find errors which correspond to the same general point in time as the visible error. When dealing with error messages for problem determination, a large software product has two conflicting requirements: Ÿ The need to dump enough information to capture any important errors Ÿ The need to make the error information easy to sift through to find a particular error

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 52% of the total text.

Page 1 of 2

Error messages with unique identifiers

   Today's complex software products require advanced Problem Determination facilities such as trace and error/informational logging. These facilities are used to keep track of any important errors, status or informational messages which may be pertinent to the workings of the software product. These facilities are not usually well connected to any user interfaces and problem determination for visible errors often entails examining error logs or traces to find errors which correspond to the same general point in time as the visible error.

When dealing with error messages for problem determination, a large software product has two conflicting requirements:
y The need to dump enough information to capture any important errors
y The need to make the error information easy to sift through to find a particular error

Consider a large software product such as a database product that has hundreds or thousands of clients using the database at the same time. If one of the clients receives an unexpected error from the database server, how can a DBA find the appropriate error in a large error log file? This can be critical for problem determination since the cause of the client's error code originated on the database server. Only the database server error logs will show the actual cause of the error and there could be a lot of error logs to sift through to find the corresponding error.

Enhancing problem determination features for a large software product is crucial since problem determination costs can become significant portion of the overall warrantee cost. Any technique which improves First Failure Data Capture or FFDC can help to reduce problem determination costs by preventing the need to recreate a problem. If the DBA is not sure which error on the server corresponds to the error that the client received, i...