一种虚拟试验系统中间件数据分发服务的设计

摘要

针对某分布式虚拟试验系统中间件的批量数据交换需求,设计数据分发服务.采用数据城实现批量数据的管理,支持对多节点多SDO属性的自由组合;采用发布订购机制在数据域和SDO属性间建立交互关系,运行时根据交互关系将SDO属性的更新同步到数据域.为进一步提高数据交换的效率,采用数据过滤机制屏蔽无效数据,减轻应用层的数据处理压力,设计C++结合Python的方式进行数据过滤,相比单纯使用C++实现更为简单、灵活和安全.服务基于ACE开发,实际测试表明服务工作正常,能满足批量数据交换的需求.%For the requirements of bulk data exchange for a distributed virtual test system middleware, data distribution service was designed. The data field was adopted to realize the management of mass data to support the free combination of multi-node SDO attribution. The publish/subscribe mechanism was employed to establish interaction relationship between the data field and SDO attribution. According to the interaction relationship, the update of SDO attribution are synchronized to the data field. To further improve the efficiency of data exchange, data filtering mechanism is used to shield the invalid data and mitigate the data processing pressure in application layer. The mode that combines C+ + with Python is designed to implement data filtering. It is more simple, more flexible and safer in comparison with the mode purely using the C+ + . The service development is based on ACE. The actual tests show that the service works well and meets the needs of bulk data exchange.