Distributed Information System for Processing and Storage of Meteorological Data

Abstract

The following article features the results of developing distributed network storage of ground meteorological observation data. The data is represented with the national variant of international rapid transmission code of environment data from meteorological stations across the Russian Federation. They are available for researchers in both visual and common export formats. The design of the distributed network storage of meteorological data includes the following modules: dispatcher module (monitors calculation nodes, distributes data to nodes, processes client requests), client module (allows external researchers to access the meteorological data), terminal module (used to import new meteorological data), data processing and storage module (a node for distributed meteorological data storage, consists of two sub-modules for data processing and data storage respectively). The article displays the results of practical testing of the developed software. To simulate the cluster of informational and calculation servers in the pilot project, multithreading was used. Multithreading is supported by nearly every operational system for parallel data processing. The development tools chosen for the network storage allowed to design storage module interaction with the optimal efficiency, to ensure proper performance, stability and reliability of processing and managing large amounts of data. The obtained results allow using the designs for efficient management of meteorological surface observation data, for rapid data gathering, for systematization and storage of hydro-meteorological data in different alphanumeric codes and other related categories.