Self-Managing Performance in Application Servers – Modelling and Data Architecture

Abstract

High performance is always a desired objective in computing systems. Managing performance through manual intervention is a well-known and obvious mechanism. The attempts to self-manage performance with minimal human intervention are predominant in the recent advances of research. Control Systems theory is playing a significant role in building such intelligent and autonomic computing systems. We are investigating in building Intelligent Application Servers by enabling control system as a first class feature in component software, at design time and runtime. In this direction, it is important to build efficient data access mechanisms that capture the control system models, performance data, analyze the data efficiently, identify patterns and build a knowledge base. In this paper we propose a data organization and architecture as a building block of developing Intelligent Application Servers.