云平台下实验室数据库资源负载优化控制仿真

Simulation of Load Optimal Control of Laboratory Database Resources under Cloud Platform

对云平台下实验室数据库资源负载进行优化控制,可提高实验室数据的管理效率。进行资源控制时,应将实验室数据库资源损耗特征按性能进行分类后,通过提高实验室数据库资源的吞吐量和整体稳定性完成控制,但是传统方法通过控制输入给数据库的负载数量来提高数据库系统的稳定,但是忽略了实验室数据库资源的吞吐量和整体稳定性,降低了数据库资源负载控制的性能。提出一种云平台下实验室数据库资源负载优化控制方法。首先,运用基于多目标的优化算法以数据库负载特征参数为负载控制的依据,根据各个负载具有的不同属性特征,获取每个负载控制时间距离内价值大且平均响应时间少的负载,确定数据库负载控制的形态特征。其次利用改进特征向量增量聚类算法,依据负载控制的形态特征对负载进行分类,采用动态聚类形式提取负载状态特征,将具备同类性能特征和数据库资源耗损特征的负载分为一类,通过负载控制以提高数据库系统的吞吐量和整体稳定性,最终实现云平台下实验室数据库资源负载优化控制。仿真结果表明,通过确定负载形态特征,然后对形态特征进行分类后实现了数据库的负载控制,优化了云平台虚拟实验室数据库资源负载的控制性能。

A load optimization control method of laboratory database resources under cloud platform is proposed. Firstly, the load feature parameter of database is used as the load control basis applying the optimization algorithm based on the multi-objective. The load with large value and short average response time in the time distance of each load control is obtained to confirm the morphological characteristics of database load control. Then, the modified in- crement cluster algorithm of feature vector is used to classify the load according to the morphological characteristics of load control, and the dynamic clustering form is used to extract the load state characteristics. Moreover, the load with same performance characteristic and loss feature of database resources is classified into one category, and the through- put and global stability of database system are improved through load control. Finally, the load optimization control of laboratory database resources under the cloud platform is achieved. The simulation results show that it can achieve the database load control to confirm the load morphological characteristics and then classify it. It also optimizes the con- trol performance of virtual laboratory database resource under cloud platform.