부산대학교 도서관

In mission-critical system's, any outage can severely impact the smooth functioning of the system. In Service Oriented Architecture (SOA)-based system, services are composed of multiple available services providing different functionalities. During the maintenance period or service upgradation, some services can become unavailable for a certain period of time. This process is known as a scheduled outage which has to be managed during maintenance or upgradation of the service. In order to effectively manage the scheduled outage of service, this study proposes Complex Service Scheduled Outage Model (CSSOM) in terms of the availability and unavailability of services in SOA-based system's. The model can assist in predicting- the dépendent services that are UP/Down during the outage period. This will help in early planning the outage of various services by taking appropriate mitigation action. Services unavailability/ availability cases in CSSOM are represented using rule-based technique in Machine Learning.