Design And Implementation Of WEB Services Supporting Distributed Timing Task Scheduling

With the rapid development of information technology and the popularization of Internet related technologies,the regular tasks have become the basis for supporting the normal operation of various businesses.The business logic of the scheduled tasks in each project is very different,but the process of scheduled tasks should be part of the business logic has nothing to do,you can put this part alone as a Web service design and implementation.At present,the existing scheduled task scheduling service lacks the support of distributed,parallel scheduling,unified management of task configuration and task scheduling,and the task itself is highly coupled.In order to solve the above problems,this dissertation starts with the issue of distributed transaction and data consistency,summarizes the basic principles of distributed technology and Paxos algorithm,and introduces the concept and purpose of ZooKeeper in detail.In order to avoid introducing too much about data consistency that is unrelated to the research topic,this paper uses the metadata of ZooKeeper cluster storage scheduling system,including the timing task configuration information and server node information;next,the remote proxy related Technology is studied,and the application of Dubbo framework in this paper is introduced emphatically.That is to say,task scheduling is decoupled from task itself.In addition,this paper introduces the related concepts of Quartz and points out the parts that need to be expanded.After that,Distributed load balancing task of regular tasks,comparative analysis of the random method,polling method,dynamic adaptive method and other algorithms,and then from the optimization point of view of a new reduction,and based on genetic algorithm designed a Load Balancing Algorithm Suitable for Distributed Timing Tasks.Through the above research and analysis,this paper completed the system design and development work,the main contents include the design of the system framework,the design and development of functional modules in each part,the design of load balancing algorithm and Quartz expansion.Finally,the theoretical analysis of the load balancing algorithm designed in this paper is carried out.The experimental results show that this algorithm can achieve load balancing of all the server nodes in the future.The Web Services are tested for their functions.According to the experimental results,the expected system Design goals.