An effective method of optimizing geospatial metadata retrieval

Nowadays, many spatial data infrastructures are providing users the ability to discover, access, publish, and share geospatial data for free. Metadata plays a critical role when searching remote sensing data entities. However, as the spatial data infrastructure usually has a large number of users, the metadata database has high access concurrency requirement. Moreover, metadata of massive remote sensing data is relatively a big table and the query conditions are complex, which includes temporal and spatial aspects. It is difficult to manage it and the metadata retrieval efficiency will not be ideal. In this paper, database cluster and table partitioning technology are adopted. Database cluster technology is used to satisfy the high concurrency requirement, and table partitioning technology is used to improve the query efficiency. Moreover, the relationship between the number of the partitions and the data volume is also studied. The results of the experiments show that all these technology greatly optimize the geospatial metadata retrieval process.