Research Of The Storage Of Massive Astronomic Star Catalog Data Based On NoSQL Database

With the development of Astronomical Observations recently, Astronomers are offered with larger and larger amount of scientific data. Among them, star catalog data are known as the meta data in Astronomic, which cause a problem of how to store and utilize them efficiently. Based on NoSQL databases, a high performance star catalog storage system is designed in this paper, in which the significant sections of cross-identification process and star catalog database are also studied and optimized.After researching the utilization of star catalog data, a star catalog storage system is proposed with the functions of cross-identification and star catalog database inquiry.Hierarchically designed, this system is composed of the Interface Layer, Middle Layer,Operation Layer and Data Layer. All these layers are independent from each other,and only communicate via packaged interfaces. Interface Layer offers user-friendly outside interfaces.Cross-identification is a fundamental means in astronomical researches of data integration. As in real-time cross-identification the datum catalog is frequently queried,it's stored in the memory database Redis in this paper, and indexed by Healpix with the help of sorted set. Compared with the practically used cross-identification implementation with MySQL, this design turn out to be obviously more efficient after the data scale up more than ten million stars.The star catalog database is a tool that is often used by astronomers in daily studies. But currently, most relational database based implementations are not well-performanced enough in data insertion and inquiry. The most common utilization of star catalog databases is space range queries(ra, dec, radius), which can be conversed to single column range queries by Healpix. In this paper, the blocks of the sky from Healpix are mapped to the chunk of distributed MongoDB. With redundant boarder data, the data queried can be performed within specific data chunks.Through plenty of experiments, the improved star catalog database proposed here,Catalog-MongoDB, is remarkably efficient compared to MySQL and reaches twice the speed of pure MongoDB in queries.