基于和声算法异构Hadoop集群资源分配优化

异构Map-Reduce环境中资源分配策略直接影响其响应时间，如何利用有效的策略将计算任务分配到计算资源是亟待解决的问题。利用和声搜索算法对异构Hadoop集群中的计算资源分配问题进行优化。对问题进行建模时考虑了异构计算机集群中各节点的处理能力、带宽和线路质量和源数据位置等因素对计算资源分配的影响，利用和声搜索算法优化资源分配策略，以期在满足用户需求的前提下提高系统的响应时间。并用Gridsim对算法进行仿真实验，实验结果表明利用和声搜索算法可以达到减少系统响应时间的目的。

Computing resource optimization in heterogeneous Hadoop cluster based on harmony search algorithm

In heterogeneous Map-Reduce environment, resource allocation strategies directly affect their response time. How to use effective strategies for the computing tasks assigned to the computing resources is an urgent problem to be solved. This paper uses harmony search algorithm to optimize computing resources in the heterogeneous Hadoop cluster environment. Taking into account the processing power, bandwidth, network quality and other factors impact on the resources allocation in the heterogeneous cloud computing environment, it models the issue and uses harmony search algorithm to optimize resource allocation strategies in order to reduce response time. It does experiments in Gridsim environment. The results show that using harmony search algorithm to optimize resource allocation can reduce response time.