Browse Prior Art Database

Error log automatic collection when encountering a specified error

IP.com Disclosure Number: IPCOM000185225D
Original Publication Date: 2009-Jul-16
Included in the Prior Art Database: 2009-Jul-16
Document File: 1 page(s) / 11K

Publishing Venue

IBM

Abstract

Instrumentation of an error log tool to make it able to recognize that an error has occurred and to collect the needed set of traces to provide support team with required documentation

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

Page 1 of 1

Error log automatic collection when encountering a specified error

    When applications are server based, that is always running, it is hard to collect traces or error logs referring to a specified error scenario in case this is not related to a blocking symptom. In other words, when an error happens and this is resulting in a wrong behavior, support teams very often ask for a collection of traces and logs to examine application execution during the failing scenario.

In the case where this scenario random occurs or simply happens from time to time, setting traces to a detailed level may result in producing a lot of useless data. Mechanisms were created to limit traces in size but even if people succeed in collecting them, it is not easy to find the exact sequence of step in tons of traces.

    There are two standard ways today to store traces and errors. The first way is to provide a location (directory or file) where traces and errors are stored. Errors and traces could be written in separated files or records but no rule is provided to delete old ones. This way should catch the error in theory since no trace is deleted. Conversely, the involved file system might also be filled without ever reaching the error condition and subsequent failing information might be lost.

The second way is to make the logger able to circularly fill the provided location. That is, when the limit of dedicated size is reached, the logger starts to write the provided location from the beginning. If this is the chosen method, an intensive production of traces might cause the info of the failing behavior to be overridden and the lost of useful traces for the analysis m...