Читать онлайн

The article discusses the problem of estimating the reliability of distributed data storage systems. The author concludes that classical models do not consider many of the capabilities of modern distributed systems, such as the ability of automatic data recovery on the rest of nodes. Such ability depends on several factors, including free and used space on nodes. This process means that such a system can recover nodes loss independently without the interaction with other agents, as system administrators, which might be required to configure new nodes for data recovery. After that, the author describes the main components of such a system, which reliability should be calculated. The parameters which are used in such an automatic data recovery are defined and added to the well-known methodology-based reliability estimating model for RAID. At the end, modelling of various data storage systems is carried out and the result of reliability estimation of these systems is compared with the systems that are based on the proposed automatic data recovery model.

Ключевые фразы: распределенные системы, надежность, ВОССТАНОВЛЕНИЕ ДАННЫХ, raid
Автор (ы): Хаммуд Обадах Рамадан
Журнал: ИСКУССТВЕННЫЕ ОБЩЕСТВА

Предпросмотр статьи

Идентификаторы и классификаторы

УДК
004. Информационные технологии. Компьютерные технологии. Теория вычислительных машин и систем
Для цитирования:
ХАММУД О. Р. MODELING THE RELIABILITY OF DISTRIBUTED DATA STORAGE SYSTEMS // ИСКУССТВЕННЫЕ ОБЩЕСТВА. 2024. Т. 19 № 2
Текстовый фрагмент статьи