Curtailing the cost of virtual machine migrations in cloud data centers
Malayam Parambath Gilesh · 2018
Each application in the cloud has its own load behavior, with independent load profile, load level, and load duration. Deployment of an application in cloud can be represented as a set of virtual machines (VM) communicating to each other. A VM may be migrated from its current host to another in various contexts such as evacuating a host machine for maintenance, balancing the workloads on host machines, optimizing physical resource utilization, and meeting the custom demands of the user applications. These migrations are costly in terms of its duration, which depends on the parameters such as the memory size of the VM, the available network bandwidth, the rate of page dirtying, and the size of the writable working set [1]. In this paper, we investigate a proactive solution to the problem. We introduce the concept of opportunistic migration of virtual machines for proactive resource management in cloud data centers. The idea is to decide whether a point in time is appropriate, in terms of the duration of a migration.