其他知识库

明安图射电频谱日像仪每天产生海量观测数据,传统的关系型数据库存储和管理这些数据时,面临着读写延迟高、性能和容量扩展能力有限以及可用性弱等诸多问题。针对这些问题,开展了基于No SQL的海量数据存储与检索应用研究。首先,详细分析了明安图射电频谱日像仪数据特点、存储需要以及面临的问题;然后,对明安图射电频谱日像仪进行数据建模,给出了明安图射电频谱日像仪的列式非关系型数据模型,同时提出了元数据和数据在No SQL中的同步存储方法,解决了二者的一致性问题,在此基础上实现了基于Cassandra的海量天文数据存储管理系统(MBDMS),最后通过实验验证了系统存储与检索的高效性、扩展性以及可行性,从结果来看,MBDMS可以很好的满足数据管理需要,是解决当前数据存储问题的一种有效方案。

Mingantu Ultrawide Spectral Radioheliograph produces massive observational data every day. In order to solve the problem of high latency of reading and writing, limited
performance of capacity expansion and weak usability when the data is stored and managed by traditional relational database, The paper has carried out research on the application of MUSER data storage based on NoSQL. First of all, deeply analyze the characteristics of MUSER data, the requirement of storage and problems being solved in detail. Secondly, the MUSER data model is given as well as the MUSER column non-relational data model. The metadata and data are synchronized in NoSQL based on Cassandra. The MUSER massive astronomical data storage management system (MBDMS) is realized based on Cassandra. The experiment validates the efficiency, expansibility and feasibility of data storage and retrieval of the system. As aresult, 
MBDMS can well meet MUSER data management needs, is an effective program to solve the current MUSER data storage problem.