A service relevance based virtual machine migration strategy in cloud

Load balancing can effectively balance the system load, improving the speed of processing and increasing the utilization of resources in cloud computing. Thus how to optimize load balancing has become one of the hot spots of current researches. The existing virtual machine migration strategies to a certain extent promoted the load balancing, but did not take into account the existence of service relationships that led to the close contacts between user avatar and the virtual machines or the links between virtual machines. It may bring more communication overhead between servers in the process of migration and computing. To this end, a virtual machine migration strategy based on service dependency is proposed. This strategy first establishes a load migration group based on service dependencies and uses it as a migration unit for virtual machine migration, and then it divides the virtual machine migration process into load balancing request phase and virtual machine migration phase for detailed design. Finally, CloudSim simulation results show that virtual machine migration strategy based on service relevance can largely reduce the communication overhead between servers while balancing the load compared with other strategies.