Collecting Data from Running Systems

Abstract

This article describes a way to build a consolidated log that can be used to facilitate the state of Windows services at runtime. It makes a side by side introduction of code instrumentation techniques and autonomic techniques that can be used to collected data from running systems in order to construct a consolidated log. The techniques try to deal with the non-standardized syntaxes and contents that come with heterogeneous logs. The collected data might be a foundation for a replacement to such logs hence the behavior can be determined by such collected data. These behaviors can be mapped to common, standardized and easy-understandable high level events that can be used as a log themselves.