Elastic resource scaling is a key feature of cloud computing. And Scale-out is the most popular approach to managing elasticity. However, this approach sometime incurs performance problems, especially for service with rapid load change. So we proposed a new auto-scaling mechanism which quickly acquires resource by Scale-up and releases resource by Scale-in without stop. In this paper, we present the design and implementation, and then we evaluated the effect of the mechanism by using workloads based on real access history. Finally we confirmed the mechanism can improve service level.
J-STAGEがリニューアルされました! https://www.jstage.jst.go.jp/browse/-char/ja/
