Reducing the energy consumption of a data center is a major goal in today’s digital world. We consider a cloud system represented by a set of physical servers hosting several Virtual Machines, and each one running tasks. We define a resource management policy that implements task migrations between overused servers to unused servers. The advantage of the policy is to balance the load and reduce the energy consumption. We model the system by a multi-server Jackson network, where each station represents a physical server, and the Virtual Machines are the servers. We derive an analytic formula for the energy consumption of the physical servers. We provide an upper bound of the migration energy for task migrations to reduce energy consumption. Moreover, we optimize the energy consumption, and we compute the migration rate that minimizes energy consumption. An extension to N>2 physical servers is presented in [2].
- Youssef AIT EL MAHJOUB
- Jean-Michel Fourneau
- Hind Castel-Taleb
- The original documents are [1, 2].
[1] "Analysis of energy consumption in cloud center with tasks migrations", Youssef AIT EL MAHJOUB, Hind CASTEL-TALEB and Jean-Michel FOURNEAU. In book, CN-Computer Networks, 2019.
[2] "Performance evaluation of Green IT networks", Youssef AIT EL MAHJOUB. In Phd-thesis, Chapter 6. link : hal.science, 2021.