摘要： 提出一种集群资源模糊聚类划分模型。对计算机集群中计算节点的CPU、内存、网络、I/O和网卡资源参数进行量化和规范化，运用模糊聚类技术，实现计算节点的聚类划分。引入任务资源需求向量和最低误差容忍向量，将计算机集群划分为若干个性能均衡的逻辑子群。测试结果表明，该模型能有效划分计算机集群，适用于云计算领域的资源调度。

关键词: 计算机集群资源, 模糊聚类, 云计算, 划分模型

Abstract: A fuzzy clustering partition model of cluster resource is proposed in this paper. It quantizes and normalizes the computer resource parameters of CPU, memory, I/O, network adapter and net. It uses fuzzy clustering technique to realize the partition of the computing nodes in the computer clusters. By using of the vector of resource demand and the vector of lowest inaccuracy tolerance, it can divide the computer cluster into several classes and the performance of these computers in one class is more similar. Test results show that this model can effectively partition the computer cluster and it fits the resource schedule of cloud computing.

Key words: computer cluster resource, fuzzy clustering, cloud computing, partition model

中图分类号: 

• TP393