Page 12 - libroBrevetti

Department of Engineering, ICT and Technologies for Energy and Transport Patent Title Consolidation od Virtual Machines in Data System for energy saving in company data centers. Centers: procedures for assignment and migration of VMs Ref. CNR 10202 Assignee(s): CNR, Università della Calabria CNR InƐƟƚute: ICAR Main Inventor: Carlo Mastroianni Countries: IT, EP, US, CA, UK, DE Priority date: 10/08/2011 Abstract The inveŶƟŽn refers to a method for assigning the computĂƟonal load to the diīerent servers of a data center by means of appropriate assignment of Virtual Machines to physical servers. The inveŶƟŽn consolidates the workload on the minimum number of servers, dynamically mŝŐƌĂƟŶŐ the Virtual Machines when needed, and puts the unloaded servers in a low power state, thus reducing energy consƵŵƉƟŽn and carbon emissions, maintaining at the same Ɵme an appropriate service level for the clients of the data center. Background The inveŶƟon addresses the problem of the huge energy consumƉƟon in the data centers, for feeding both servers themselves and cooling systems. It has been esƟŵĂƚĞd that the consumed energy for the IT infrastructures are equal to about 3% of the total produced electric energy. The energy consumƉƟŽn impacts both on the company’s costs owning the data centers and on their clients, and has also a great environmental eīect due to the CO2 emissions. Technology The invenƟon is based on an innovaƟve bio-inspired probabilisƟc algorithm which consolidates the maximum number of virtual machines on the minimum number of physical servers in a data center, enabling the switch ŽīͬhibeƌŶĂƟŽŶ of those freed-up, making them dynamically available as addiƟonal capacity for incremental workloads. Advantages and ApplicaƟŽŶs The inveŶƟŽn ŽīĞƌƐ a ƉƌĂĐƟĐĂl soůƵƟŽn to a problem that is diĸcult to solve with classical centralized and determinisƟĐ algorithms. Advantages are crucially relevant for both business as well as technical audiences, and go far beyond energy saving alone: monitoring, capacity planning, orchesƚƌĂƟon opƟminjĂƟon, risk monitorŝŶŐͬmŝƟŐĂƟŽn. The involved market is huge and growing, with no established alterŶĂƟve soůƵƟŽŶƐ͘ Development stage The inveŶƟŽn is the core of the sŽŌware product implemented and commercialized by Eco4Cloud, a CNR spin-oī based in Rende (CS). The Eco4Cloud soŌware has been licensed to Telecom Italia for 3 years, stĂƌƟng from January 2014. The product is being commercialized on the iŶƚĞƌŶĂƟonal market of data centers, stĂƌƟng from UK. 3