Usage pattern based survivability prediction of web system
Xun Lee, Jiaan Zhou, Hao Fu · 2016
With the rapid development of SOA technology, the instability of the distributed Web services has become a hot research topic. The survivability of Web system at a certain period of time is an important and useful indicator for making an appropriate adjustment dynamically to ensure the system runs best while the runtime environment or requirements are changed. However, most researchers only focus on estimating the overall survivability of Web system. The overall survivability cannot reflect the QoS of Web system under different usage patterns actually. In this paper, we propose an approach based on Web log statistic to predict the particular survivability of Web system under a particular usage pattern.