The Design And Implementation Of A Safe City System’s High-Availability Mechanism

Safe city system is an iot system which is developed by a research institute. The system is composed of many business systems. These subsystems are used to implement the integration management function, which includes urban security, traffic scheduling and emergency medical treatment scheduling, etc. Safe city system is divided into business systems layer and resource management layer. The business systems layer includes variety business subsystems. These systems are complete systems. Each of them is responsible for implementing a business function, such as traffic scheduling. At the same time, these systems work together for the integration scheduling function. The resource management layer uses Infrastructure as a Service technology to manage the underlying physical resources and provide the resources in the form of a virtual machine for the upper business systems.Safe city system is a complex integrated system. The size of the system is very big. In order to make the system be able to provide stable and reliable service, we need to deploy high availability mechanism for the system. By using high availability mechanism, we can reduce the probability of failure, shorten maintenance time and increase the availability of the system.In order to improve the usability of the system, on the one hand, we need to enhance the reliability of the business system. On the other hand, we need to increase the availability of the virtual machine. This article mainly discusses the high availability mechanism of the virtual machine.The system of the resources management layer is based on an open source platform which is named CloudStack. Because the size of the business system is very big, the virtual machine failure rate is high, and the maintenance cycle is very long, the current system can’t satisfy the requirement of the system availability. Based on the research of existing problems, we find that exceptions occur easily when the virtual machine is in transition state. When an exception occurs, the virtual machine will be in abnormal state. The resources will not be released until the exception has been resolved by the administrator. The business system on the virtual machine will be down, too.Considering the existing problems of the system, we design the high availability system mainly from two aspects. Firstly, we need to strengthen the function of handling the exception of the virtual machine. The system will monitor the state of the virtual machine in real-time. When an exception occurs, the system will capture the exception and analyze the type of the exception. According the analysis, the system will handle the exception. Secondly, we will add the function of migration to the virtual machine. The more important virtual machines, which need add a tag, will have the function. When the virtual machine is down, the virtual machine will restart in its previous host, or another new available host. Thus, we can decrease the time of outage. Through these two approaches, we can effectively shorten the processing time, avoid the virtual machine being in abnormal state for a long time and make the virtual machine available as soon as possible. Through the deployment of high availability mechanism, we can improve the reliability of virtual machine, so as to improve the usability of the system.In this paper, based on the basic concept, principle and method mechanism of the high availability mechanism, related technologies are introduced. Then we analyze the safe city system demand and technical requirements of the application of high availability mechanism, and respectively model the system of data and process. We illustrate the system architecture design of application, the design of the system specific business process. Finally we introduce the realization of the system high availability system.Based on the realization of the mechanism of the high availability of safe city system, the test results show that the high availability system can better improve the availability of system, which can make the system more stable running, effectively improve the reliability of the system as a whole and achieve the target of design.