Measuring Load of Complex Service-oriented Systems

Abstract

Complex software systems produce a large amount of data depicting their internal state and activities. The data can be monitored to make estimations and predictions of the status of the system, helping taking preventative actions in case of impending malfunctions and failures. However, a complex system may reveal thousands of internal metrics, which makes it a non-trivial task to decide which metrics are the most important to monitor. This thesis uses a performance test bench tool to generate load of different intensities on the target system, which is a specific service-oriented application platform. The numeric metrics data collected from the system is combined with the load intensity at each moment. The combined data is used to analyse which metrics are best at estimating the load of the system. The analysis part consists of preprocessing and two main analysis tasks: regression analysis and cluster analysis. Regression analysis rates and ranks the metrics by their ability to measure the load of the system, while cluster analysis groups similar and mutually redundant metrics together. The combined results form a concise list of groups of best metrics to follow in the system. The results show that the most important metrics are related to network traffic and request counts, as well as memory usage and disk activity. The results help with the designs of efficient monitoring views and powerful machine learning prediction models. In addition, the thesis makes suggestions for improving the process of the conducted study.