Тип публикации: доклад, тезисы доклада, статья из сборника материалов конференций
Конференция: International Scientific and Research Conference on Topical Issues in Aeronautics and Astronautics, TIAA 2016; Красноярск, Россия; Красноярск, Россия
Год издания: 2016
Идентификатор DOI: 10.1088/1757-899X/155/1/012041
Ключевые слова: Availability, Closed loop control systems, Cluster computing, Communication satellites, Disasters, Fault tolerance, Fault tolerant computer systems, Satellites, Space flight, Cluster architecture, Communications satellite, Critical functions, Distributed control, Distributed systems, Information processing center, Information processing systems, System availability, Distributed parameter control systems
Аннотация: Control and information processing systems, which often executes critical functions, must satisfy requirements not only of fault tolerance, but also of disaster tolerance. Cluster architecture is reasonable to be applied to provide disaster tolerance of these systems. In this case clusters are separate control and information proceПоказать полностьюssing centers united by means of communication channels. Thus, clusters are a single hardware resource interacting with each other to achieve system objectives. Remote cluster positioning allows ensuring system availability and disaster tolerance even in case of some units' failures or a whole cluster crash. A technique for evaluation of availability of cluster distributed systems for control and information processing based on a cluster quorum is presented in the paper. This technique can be applied to different cluster distributed control and information processing systems, claimed to be based on the disaster tolerance principles. In the article we discuss a communications satellite system as an example of a cluster distributed disaster tolerant control and information processing system. Evaluation of availability of the communications satellite system is provided. Possible scenarios of communications satellite system cluster-based components failures were analyzed. The analysis made it possible to choose the best way to implement the cluster structure for a distributed control and information processing system. © Published under licence by IOP Publishing Ltd.
Журнал: IOP Conference Series: Materials Science and Engineering
Выпуск журнала: Vol. 155, Is. 1