Design Of Monitoring Server System For Landslide

Landslides are the most frequent of all geological disasters and cause the most losses.Internet technology is used to remotely monitor the state of mountain slopes,and early warning information is issued to user groups to achieve data prediction and information sharing,which effectively reduces people 's lives and the loss of property has important practical significance.In view of the fact that most of the current monitoring methods belong to the monitoring of single slopes,the monitoring results are only used as a simple basis for whether to protect the mountain,and the information is not communicated to the people facing danger,and the problem of real-time warning cannot be real-time.The system realizes the prediction and early warning of landslides.The system adopts the B / S architecture and separates the front and back ends to realize user management,role management,equipment management,data management,mountain management and other functions.By collecting,transmitting,analyzing and processing mountain data,real-time warning of the state of the mountain is realized,And visual display of data.In order to solve the pressure of concurrent requests on a single server and speed up the system response,the paper constructs an LVS load balancing cluster and proposes the following innovations.A LVS dynamic load balancing algorithm based on PI negative feedback is designed.By calculating the weight of server performance impact factor,the weight value to be allocated to each node is dynamically set,so as to better achieve overall server load balancing.The algorithm periodically collects server performance impact factors,including CPU utilization,memory occupancy,network bandwidth utilization,and I / O utilization,determines the weight of factors according to the rate of change of each factor,and calculates the real-time load of the current node,Using the incremental idea of PI control,dynamically adjust the weight of the server node according to the real-time feedback value of the server load.Experiments show that the algorithm improves the system's response speed by 20%-25%,can effectively loadscheduling different requests,and improves the load capacity of the cluster system.In order to further improve the response speed of the system and reduce the read and write pressure of the database,the paper implements a cache layer on the database,and designs a cache strategy based on the data characteristics,analyzes the size of the cache block and updates and replacements,increases the cache hit rate and reduces Cache update frequency.The landslide monitoring server system designed by the paper can identify and process sensor data and mountain images,and manage the monitoring equipment to realize the forecast and early warning and information release of landslides.Sex.The experimental results show that the design of the entire system meets the expected requirements in terms of accuracy,realtime,and effectiveness,and it has certain reference significance for the study of the entire landslide monitoring.