Energy Efficient Virtual Machine Migration Algorithm for Virtual Machine Placement in Cloud Data Centers for Green Computing

Abstract

Cloud technology also offers other benefits to developing countries since they newlineno longer have the burden of investing in costly infrastructures and can tap into data newlineand applications that are readily available in the cloud. While there may be challenges newlinein integrating old or current tools with new technologies and the corresponding level of newlineservices, the benefits will outweigh the inhibition to move to the cloud. newlineThis new catalyst in the IT world, cloud computing, has traditionalists spinning newlineand adventurists very excited. The effect of cloud computing on traditional data centers newlineand infrastructures will not be trivial. Virtualization is a technique pervasively applied newlinein modern datacenters (DCs) to maximize resource utilization, reduce greenhouse gas newlineemissions, and lower overall cost. Virtual machine (VM) migration is widely exploited newlinewithin and across DCs to meet a variety of needs of the virtualized cloud environment. newlineFor instance, server consolidation needs VM migration for power management. newlineThe VM migration process is very resource intensive and requires intelligent newlineapproaches to avoid saturating network bandwidth and to keep server downtime to a newlineminimum. Inefficient resource allocation makes DCs underutilized, increasing newlineelectricity consumption, carbon dioxide emissions and climate changes as well. To newlineprevent resource wastage due to low utilization, server consolidation is efficiently newlineapplied by encouraging VM migration to pack servers workload into minimum newlinenumber of physical machines. It leads a situation where under loaded servers may be newlineENERGY EFFICIENT VIRTUAL MACHINE (VM) MIGRATION ALGORITHM FOR VIRTUAL ANIL DIXIT newlineMACHINE PLACEMENT IN CLOUD DATA CENTERS FOR GREEN COMPUTING newlineswitched to hibernate mode after the VMs in it are migrated, and this paves the way for newlinereducing electricity consumption and resource wastage. newlineServer consolidation technique has to deal with four scenarios: 1. Deciding if a newlinehost is considered to be under-loaded, so that all VMs should be migrated from it, and newlinethe host should be switched to a low-power mode. 2. Deciding if a host is considered to newlinebe overloaded, so that some VMs should be migrated from it to other active or newlinereactivated hosts to avoid violating the QoS requirements. 3. Selecting VMs to migrate newlinefrom an overloaded host. 4. Placing VMs selected for migration on other active or newlinereactivated hosts. With integration of cloud in technologies such as Internet of Things newline(IoT) , BigData management, Wireless Body Area Network (WBAN), Vehicular Adhoc Network (VANET) and etc. there exists various conflicting requirements in both newlinecustomer and provider perspectives to be met. For instance, VM migration happens newlineover inter/intra network links for load balancing to offload overloaded servers into newlineunderutilized servers while it must meet the need of QoS-intensive applications. newlineHence, smart approaches in VM Selection and Placement will take over the newlineproblem since untimely/aggressive VM migrations within/across DCs newlinecause performance degradation and saturate links of Local Area or Wide Area Network newlineboundaries. VM selection and VM placement algorithms need to be designed newlineseparately so that a close to optimal migration can take place to ensure minimum newlinenumber of active physical server and minimum violation of Service Level Agreement newline(SLA) resulting in maximum resource utilization and minimum waste. newlineThis work proposes to develop energy efficient VM selection and placement newlinealgorithms to achieve optimal VM migration. Four new VM Selection policies and newlinethree new VM Placement policies are proposed. The effectiveness of these proposed newlinealgorithms is verified through simulation with CloudSim tool. newlineThe objectives of this work are to save energy during VM migration, reduce the newlinenumber of migrations and host shut downs, with the ultimate aim of increasing resource newlineutilization and simultaneously reducing the power consumption so that the entire VM newlinemigration process becomes more effective and efficient.