面向工业控制云服务优化的运行时热点数据及其收集系统研究

Research on Runtime Hotspot Data and Its Collection System for Industrial Control Cloud Service Optimization

承载工业控制系统云服务的计算机及其上运行的系统软件在平台即服务(PaaS)的思想指导下被抽象成为服务。为提高计算效率,降低硬件及运维成本,需要云服务商、芯片设计者、用户共同努力。为此,面向工业控制系统收集云服务商提供的主要业务种类,阐述云服务的优化策略,分析工控系统的运行时热点数据。同时开发运行时热点数据收集系统,收集业务系统调用的接口及其方法的调用频度、规律等数据,利用这些数据辅助专用计算机设计、系统软件运行优化。此外,目前的EDA工具在架构设计、版图制作、工艺选择时也十分依赖于这些运行时数据,系统收集到的热点数据能帮助芯片设计者选择最优算法固化在芯片中。

The computer that carries the cloud service of the industrial control system and the system software running on it are abstracted into a service under the guidance of the platform as a service(PaaS).In order to improve computing efficiency and reduce hardware and operation maintenance costs,it requires the joint efforts of cloud service providers,chip designers and users.Therefore,for industrial control systems,the main business types provided by cloud service providers are collected,the optimization strategy of cloud services is expounded,the run⁃time hotspot data of industrial control systems is analyzed,and a runtime hotspot data collection system is developed.The runtime hotspot data collection system collects data such as the calling frequency and regularity of the interfaces and methods invoked by the business system,and these data assists the design of special computers and the optimization of system software runtime.In addition,the current EDA tools also rely heavily on these runtime data in architecture design,layout making,and process selection.The hotspot data collected by the system can help chip designers choose the optimal algorithm to be solidified in the chip.