The Design And Implementation Of Distributed Task Scheduling And Management System For Commercial Bank

With the rapid growth of China's economy,the pace of development of the banking industry has also been significantly accelerated,the domestic addition to the Industrial and Commercial Bank of China,Agricultural Bank of China,China Construction Bank,Bank of China,Bank of Communications,the five major state-owned banks,recently established a large number of new urban business bank.Although the size of urban commercial banks is far less than the state-owned banks,but the scope of business covers more than 80% of state-owned banks,it also requires a large number of background IT systems to support operations.IT systems in the banking industry can be divided into four categories,namely,business support,business analysis,regulatory reporting and maintenance of the class.This topic is a distributed task scheduling and management system for commercial bank.The IT systems of this city commercial bank is about 200,and the scale of IT systems of medium-sized commercial banks is similar.Of these 200 IT systems,most of them belong to the first three categories.The bank will aggregate,analyze and migrate data within the IT systems after the day-to-day operation.This operation is called daily batch operations.Most of the bank's IT systems need to perform their own batch operations every day.Nowadays,the majority of bank IT system batch operations using serial execution,and gradually complete the daily batch operations.But with the continuous increase in the business,the requirements of the timeliness is also getting higher and higher.The traditional way of serial operation has been unable to meet the needs of future development.At the same time,each IT system also exists between the more closely dependent.Therefore,the system designed and implemented in this project can maximize the ability of batch operation to perform in parallel,and can simultaneously schedule the batch operation of multiple IT systems.The bank for the data security requirements are relatively high,the system with open source distributed message queue system Rabbit MQ as the task data transmission channel,to minimize the complexity of network connection among the programs,make data redundancy storage by setting policies,and also improve the horizontal expansion of the system capacity.The system designed mainly includes four functional modules,database storage and computing function module,task scheduling service module,distributed message queue module and task execution module.the task scheduling service module and the task execution module use the C / C ++ language to implement the independent background programs.The database storage and computing function module uses the SQL language to implement the related functions.The distributed message queue module mainly implements data redundancy storage and load balancing.Task scheduling server program using distributed architecture design,with master and slave mode running,can dynamically increase or decrease the service node,automatic failover operation,improve the system's parallel capacity and horizontal scalability.In addition,the task scheduling server program has been packaged with the container technology,easy to program deployment and migration.The system of this project can realize the capability of no single point of failure in database,task scheduling server and distributed message queue system.