Real-Time Performance Prediction for Cloud Components
Yilei Zhang, Zibin Zheng, Michael Rung-Tsong Lyu · 2012
Cloud computing provides access to large pools of distributed components for building high-quality applications. User-side performance of cloud components highly depends on the remote server status as well as the unpredictability of the Internet, which are variable over time. It is an important task to explore an method to predict the real-time performance of cloud components. To address this critical challenge, this paper proposes a prediction framework to predict real-time component performance effectively. Our prediction framework builds feature models based on the past usage experience of different users and employs time series analysis techniques on feature trends to make performance prediction. The results of large-scale experiments show the effectiveness and efficiency of our method.