A parallel virtual machine aggregation method based on spark and optimized mbbo algorithm

Abstract

The invention provides a Spark and MBBO (multi-objective biogeography-based optimization) algorithm-based parallel virtual machine aggregation method. Through an extended Spark parallel framework and an MBBO algorithm, an optimal migration solution of a VMC problem can be solved in a relatively short convergence time, and a foundation is laid for subsequent virtual machine parallel migration. The method comprises the following steps of 1, mapping a virtual machine aggregation problem to a biogeography-based optimization algorithm, determining constraint conditions, and clearly determining solving objectives; and 2, based on the extended Spark parallel framework, distributing an initial habitat group meeting the constraint conditions to Spark computing nodes, iteratively executing an MBBO parallel algorithm, stopping algorithm execution until a stop condition is met, and obtaining an optimal solution capable of balancing a plurality of optimization solving objectives. While the extended Spark parallel framework is utilized, a biogeography concept is mapped to an optimization problem.

Description

technical field [0001] The invention relates to the field of virtual machine aggregation in computer science and technology, specifically a parallel virtual machine aggregation method based on Spark and an optimized MBBO algorithm. Background technique [0002] As the hardware resources of the unified management data center, virtual machine VM (Virtual Machine) has reached an unprecedented scale of use, which brings two major problems, high energy consumption and unbalanced load. At present, the above two problems are mainly solved by virtual machine aggregation VMC (Virtual Machine Consolidation) technology, which optimizes energy consumption, load balancing, network bandwidth and other aspects of the data center by rearranging the placement of VMs on the host HM (Host Machine). . [0003] VMC achieves this by establishing a mathematical model and proposing an effective aggregation scheme in massive HMs and VMs with complex topologies, and dynamically migrating VMs from so...

AI Technical Summary

Problems solved by technology

The heuristic binning algorithm and dynamic programming algorithm are simple to deploy and fast to solve, but only a few dimensions of computing resources can be considered, which is not enough to solve complex virtual machine aggregation problems

Bionic algorithms such as genetic algorithm and ant colony algorithm can deal with the limitations of various resources, but the search speed is slow, and there are disadvantages of complex configuration parameters, which greatly affect the search results

Images