Method Improvement And Application Design Of Cloud Computing Resource Scheduling

Based on the service quality cloud computing resource scheduling method design,this paper puts forward higher requirements for the high-real-time,online service model of cloud computing service in view of the commercial characteristics and service model of cloud computing at home and abroad,and the traditional resources In the scheduling strategy,only the virtualization computing,memory,storage and network resources are allocated to limit the service capability of LC(Latency Critical)load,increase the service delay and reduce the quality of service,and study the emerging CAT(Cache Allocation Technology)technology,the innovative use of PID(Proportion Integration Differentiation)algorithm as a resource scheduling controller,the use of WCP motherboard as a multi-chip processor(CMPs)platform,with Intel? Xeon ? E5 series Processor,based on the Mesos resource scheduling system,proposed for multi-core,chip environment,cache resource scheduling method to make up for the industry in the cache resource control defects,the topic has a strong relevance and practicality.Based on the Linux operating system,the Mesos resource scheduling system and the Marathon framework are used to build and simulate a Mesos cluster environment with master nodes and two slave nodes.The load start process,resource allocation mode and state of the Marathon framework are tested Monitoring function.In the further study,the use of WCP motherboard and Intel? Xeon ? E5 series of processors as a hardware service platform,combined with Mesos scheduling system and Marathon framework software,through the use of C language in the Linux system to complete the PCM data monitoring module,PID control Module,front-end parameter reading module,back-end data write module design,and the use of SQL language in the Linux system to complete the Mysql database design,including front-end parameter storage and back-end data storage table structure,based on the use of HTML,JavaScript language to complete the back-end data visualization system design,carried out four back-end data dynamic display,PID control switch,PID algorithm parameter configuration,load addition and expansion function design.The research of this project has certain reference value to promote the development of cloud computing resource scheduling technology.The research work of this paper is mainly reflected in the following aspects: First,the status and latest achievements of cloud computing resource scheduling methods at home and abroad are studied.Combining the limitations of traditional resource allocation methods,CAT cache allocation technology,PID automatic control Algorithm,cloud computing resource scheduling technology,Docker container technology analysis,focusing on the cloud computing resource scheduling method in the automatic control module works and protocol standards.Second,the study of Mesos resource scheduling system and Marathon framework of the relevant technology,in the relevant platform and software on the basis of comparison and selection,in the three Linux host using Linux Shell command to install a Mesos master node,two Mesos Slave node,and in each node to install the PID controller,set up a miniature Mesos cluster,the use of Docker container technology to load the container package,the PID controller function test and performance testing,in order to master the entire Mesos The master nodes in the cluster control the communication between nodes.Third,the PID controller is designed so that the load has the function of automatically adjusting the performance data through the PID control item,and the real-time data can be sent from the PCM data monitoring item to the inside of the controller as the real data and the controller The target data is used to calculate the difference,and the control output value is obtained by the mathematical formula of the PID control algorithm,and the allocation of the cache resource is controlled according to the output value.In addition to the Mysql database design,so that the back-end data and front-end parameters to achieve data persistence,to store data from the front-end program.Fourthly,the design of the back-end data visualization system is studied.The working principle and working mode of the Web server are studied.By analyzing the visualization requirements of the PID controller,we choose the appropriate Web server and use the HTML and JavaScript language to develop the data Display,control buttons,parameter adjustment frame,load add frame design and function test.Fifth,the whole system hardware and software together to do the overall functional testing and performance testing,including the start load,set the size of the cache resources,back-end data storage,front-end data read,data dynamic display test,select a variety of LC Load and other loads for performance testing.The test results show that the whole system meets the requirements of the intended design.