Design And Implementation Of DCIM System For Small And Micro Data Center

Under the tide of Internet economy,data center has become an important infrastructure,and with the arrival of 5G era,the demand of IDC industry has shown a continuous explosion,so it has become a hot topic in the industry for how to manage data center efficiently.In order to enhance the intelligent management of data centers,various types of systems have emerged in the market,including DCIM(Data Center Infrastructure Monitoring and Management)system,which is a comprehensive monitoring and management system integrating power environment monitoring,asset capacity management and PUE energy management.However,most DCIM systems at this stage are developed for large data centers,which require multiple high-performance computers and large screen monitors,and their deployment costs are unaffordable for small and micro data centers;in terms of software functions,some of the functions of large DCIM systems are designed to achieve little effect for small and micro data centers,which undoubtedly causes a waste of resources;at the same time,most of the systems do not have a good understanding of early warning mechanisms,intelligent This paper examines the shortcomings of current DCIM systems.In this paper,we analyze the shortcomings of current DCIM systems and propose a DCIM system design scheme for small and micro data centers.The scheme takes the intelligent terminal as the core hardware platform,designs and develops the hardware circuit and management software of the intelligent terminal,and completes the research on the early warning mechanism and the development of the mobile APP.The system is characterized by low cost,high level of intelligence and remoteness,and good scalability and flexibility.The main work of this paper is as follows.(1)analyze the background of the DCIM system and the current research status and then propose the requirements and design objectives of the system,and also determine the overall architecture and topology of the system.(2)Design and implement the hardware and management software of the intelligent terminal.Using a high-performance chip as the core processor of the intelligent terminal and equipped with Android operating system,we design and implement the main board circuit,serial expansion board circuit and integrated IO expansion board circuit of the intelligent terminal.At the same time,the intelligent terminal management software is developed,the functional architecture and database of the software are designed,and the key functions such as data acquisition function,graphic interface signal connection function,power environment monitoring function,and asset capacity management function are designed in detail,and finally the development of the intelligent terminal management software is completed.This part of work solves the problems of high hardware cost and waste of software function resources of DCIM system,and realizes the infrastructure monitoring system which is more suitable for small and micro data center.(3)Research on data center temperature and humidity early warning mechanism based on neural network algorithm,use ELM extreme learning machine for temperature and humidity prediction and complete comparison with actual data;use BP neural network algorithm for environmental early warning classification and complete training and result analysis,and study system linkage control function according to early warning result.This part of the work to improve the intelligence level of the system.(4)Design and develop the mobile APP,complete the overall design,authority classification design,server scheme design,backend API interface design,application interface design and remote information push function design of the mobile APP.Finally,the test platform is built and the system hardware and software tests are completed.The results show that the system performance is stable and the functions are perfect,which verifies the feasibility of this design scheme and provides some reference for the research of DCIM system for small and micro data centers at this stage.