| With the continuous development of information technologies such as navigation and mapping,a large number of geospatial data are generated,and people’s daily life is increasingly dependent on spatial data,Therefore,the era of spatial big data has arrived.The traditional geospatial data service platform facilitates the development of services such as storage,processing and query of spatial data,but it is increasingly unable to meet the changing needs of users due to its complex platform protocols,huge volume,and shortcomings in scalability and maintainability.In order to solve the above problems,the thesis adopts Spring Cloud,Docker,Kubernetes and other core technologies to design and implement a microservice system for spatial data processing.The system realizes the main functions of interactive orchestration,service combination and microservitization of various spatial data services,and satisfies the characteristics of decoupling and decentralization of microservices,and has the advantages of agile deployment and rapid launch.The main work of the thesis is as follows:(1)Based on Spring Cloud,the thesis designs and implements a server-side software framework of microservices which contains various types of spatial data processing.The framework supports the effective management and monitoring of various microservice instances,supports the asynchronous calls of time-consuming microservices based on messaging mechanism,and supports load balancing with the characteristics of spatial data processing services.(2)On the basis of the first step,the thesis uses Kubernetes as the orchestration engine,and docker as the execution engine to build a container cloud platform,and designs and implements an interactive container orchestration method that supports rapid launch.(3)Based on the first two steps,the thesis studys the microservice composition technology based on Medley for WPS spatial data processing microservice,and designs and implements a basic framework of microservice composition,whith supports DSL of microservice composition.For the spatial data processing microservice system,the thesis builds a distributed experimental environment.Based on the GDELT data set,three functions of spatial data processing microservice system are evaluated by several experiments,such as load balancing,service orchestration and service composition,and the experimental results show that:(1)When the request has obvious different distribution characteristics,with the increase of the number of requests,the load balancing strategy in special scene performs better than the general load balancing strategy;(2)The efficiency of interactive orchestration method is significantly higher than that of traditional command line;(3)The composite service runs smoothly under requests of different durations. |
