基于内存数据库的海量数据实时处理策略

Strategy of Massive Data Real-time Processing Based on Memory Database

在大型的作战指挥信息系统中,庞大的雷达情报、指挥控制、模拟仿真等实时报文信息以及来自磁盘数据库中的数据对系统的高效、快速、及时响应提出了很高的技术要求。目前单一的磁盘数据库Oracle很难满足这种高时效、低延迟的需求。针对大型作战指挥信息系统海量数据实时处理的需求,提出一种新的基于内存数据库的海量数据实时处理框架,并与传统作战指挥系统中单一Oracle数据库数据处理进行综合比较,提出改良方法和具体设计方案,通过实验说明新框架的优势。

In large-scale operational command information system, the real-time message information such as radar intelligence, command control, simulation, etc. and the data from disk database put forward high technical requirements to the efficiency, speediness and real-time response of the sys-tem. Current single disk database Oracle is difficult to meet the demand of high time efficiency and low delay. Aiming at the request to massive data real-time processing in large-sale operation com- mand information system, this paper proposes a new mass data real-time processing framework based on memory database, and compares it with the single Oracle database data processing in tra-ditional operational command system, then puts forward the improvement method and specific de-sign scheme, shows that the new framework has many advantages through experiments.