Service monitoring is an important part in web service systems, since it aims to ensure that the functions and performance of the service have fulfill the customers' requirements, and detect the potential risks as early as possible. However, monitoring has its costs, which will cause the performance degradation of monitored service system and large resources consumption. If monitoring costs are higher than monitoring benefits, service monitoring will be meaningless. Therefore, it is necessary to balance the monitoring benefits and monitoring costs.
Efficient service monitoring means cost-efficiently and time-efficiently of service monitoring. In order to maximize the functions of service monitoring, careful study on how to improve the efficiency of service monitoring is required.
This thesis mainly presents a taxonomy of different approaches of service monitoring. It covers several existing optimization approaches, and describes them by giving examples. Since there are two main factors that will add monitoring costs, large amount of monitored objects and inappropriate monitoring approaches, this thesis makes comparisons among different monitoring technologies. They can provide options on selecting the appropriate monitored objects and monitoring approaches to save monitoring costs. Finally, this thesis describes some limitations of optimization approaches, and discusses possible extensions on improving the efficiency of service monitoring.
ACM Computing Classification System (CSS):
Applied computing → Enterprise computing → Business process management →Business process monitoring
