运载火箭试验大数据存储架构设计与应用

运载火箭试验产生的数据量呈现出爆炸式增长，主要特点表现为数据种类多、数据密度大、数据持续时间长。传统单机部署和基于关系型数据库与文件的系统架构的不足逐渐显现，不同种类的数据不做区分存储，存储和查询效率低，数据无备份，存在单点故障导致数据丢失的问题，无法满足海量数据场景下的存储计算业务需求。利用大数据技术思想，针对运载火箭存储计算业务的需求，设计出一套运载火箭试验大数据存储架构，并给出了各存储组件的存储模型设计方法。通过实际工程应用表明：该架构具备良好的可靠性、可扩展性和可维护性，是一种切实可行的大数据存储架构设计，能够满足运载火箭试验数据的存储计算等业务需求。

Architecture design and application of big data storage for launch vehicle test

The amount of data produced by launch vehicle test shows an explosive growth, which is mainly characterized by complex data types, large data density and long time duration. The system architecture shortcomings of traditional stand-alone deployment based on relational database and file are gradually emerging. Different types of data stored together, low efficiency of writing and query, no backup of data, and single point of failure could not meet the needs of storage and computing business in the context of massive launch vehicle test data field. Using the idea of big data technology and aiming at the need of launch vehicle test data storage and computing, a set of big data storage architecture for massive launch vehicle test data is proposed, and the storage model design method of each storage component is designed. The practical engineering application effect shows that the architecture has good reliability, scalability and maintainability. It is a practical big data storage architecture, which could meet the launch vehicle test data business needs of storage and calculation.