1. Analyzing live migration energy overhead in cloud computing: A survey
- Author
-
Salma K. Elbay, Islam Hegazy, and El-Sayed M. El-Horabty
- Subjects
business.industry ,Computer science ,Distributed computing ,Process (computing) ,020206 networking & telecommunications ,Cloud computing ,Workload ,02 engineering and technology ,Power (physics) ,0202 electrical engineering, electronic engineering, information engineering ,Bandwidth (computing) ,Overhead (computing) ,020201 artificial intelligence & image processing ,business ,Live migration - Abstract
Live Migration is the backbone of dynamic VM consolidation in modern data centers. However, live migration process comes with a cost, including VM performance degradation, downtime and energy cost. In this work, we present and evaluate the research attempts to quantify the energy cost that accompanies live migration process. In order to measure the energy overhead, the amount of power drawn during migration and the total migration time are needed to be quantified using mathematical models. We find that the network bandwidth and the page dirtying rate varying according to the type of workload are the major factors influencing the amount of power drawn by a physical server. While the VM size and the network bandwidth are the factors affecting the total migration time.
- Published
- 2017
- Full Text
- View/download PDF