Information system of the general observation archive of SAO RAS. Current state

Abstract

There is no doubt about the need for long-term storage of observations since they do not lose their scientific significance when studying the variability of celestial objects and other tasks. Maintaining archival systems requires software development, metadata management, fault tolerance, and periodic migration of files to modern storage media, which is important to ensure long-term safety of information. Providers of small collections need to join forces and share experiences so as not to “reinvent the wheel”. According to the standards accepted in the astronomical community, the SAO RAS observation archive belongs to small collections compared to the collections of modern astronomical data centers. Here we present the basis and methods of our archive system development. The archive includes more than two dozen digital collections with observations obtained on various observation instruments. It consists of three interconnected components that provide accumulation, long-term storage and access to data. The volume of collections is approximately 2.5 TB. The information system is implemented on the basis of the PostgreSQL software. Access to archived files is provided through a web interface that supports queries by observation instrument and date, coordinates, object name, program applicant, observer. The database contains more than 4 million records. The archive system uses two servers with directly connected storage systems. The work server implements data access, and the test server is used for development and testing. This configuration provides both permanent access to files and the ability to develop the information system.