МЕТОД НАДІЙНОГО ЗБЕРІГАННЯ ДАНИХ НА ОСНОВІ НАДЛИШКОВОЇ СИСТЕМИ ЗАЛИШКОВИХ КЛАСІВ
THE METHOD OF RELIABLE DATA STORAGE BASED ON REDUNDANT RESIDUE NUMBER SYSTEM
Сторінки: 98-104. Номер: №6, 2019 (279)
Автори:
В.В. ЯЦКІВ, С.В. КУЛИНА
Тернопільський національний економічний університет
V.V. YATSKIV, S.V. KULYNA
Ternopil National Economic University
DOI: https://www.doi.org/10.31891/2307-5732-2019-279-6-98-104
Рецензія/Peer review : 10.11.2019 р.
Надрукована/Printed : 2.01.2020 р.
Анотація мовою оригіналу
В роботі запропоновано метод надійного зберігання даних на основі коригуючих кодів системи залишкових класів. Важливою характеристикою систем зберігання даних є висока швидкодія запису та зчитування інформації з накопичувачів. Враховуючи, що система зберігання базується на коригуючих кодах, необхідно отримати підвищену швидкодію процесів кодування та декодування даних. Перевагою запропонованого рішення є менша надлишковість для відновлення даних, порівняно з технологією RAID та інших систем, що базуються на коригуючих кодах.
Ключові слова: система залишкових класів, коригуючі коди.
Розширена анотація англійською мовою
The main advantages and disadvantages of existing real-time data storage methods used to improve the reliability of storage systems are discussed. The main purpose of the study is the storage system reliability improvement by using coding methods based on redundant residue number system. The proposed method of secure data storage is based on the corrective codes of the residue number system. The essence of this method is that the array of obtained data is divided into fragments which are used to calculate the residuals by the selected system of relatively prime modules. The resulting residuals are stored on separate disks, which can be placed on one device or connected remotely. Considering that during encoding an extended module system is used, in case of disk distortion or failure, the redundant residue number system allows to recover the data fragment using only one validation module, if it is known which data block has been lost. The important characteristic of storage systems is the high speed of data writing / reading. Due to the fact that the storage system will use corrective codes, it is necessary to ensure a high speed of data encoding / decoding processes. The advantage of the proposed solution is that there is less redundancy in data recovery compared to RAID technology and systems that use other correction codes. The authors conducted experimental studies on the dependence of the required time for data packet processing on the selected module systems. According to the obtained results, the transition to a system with more modules allows to increase the speed by 13-15% compared to the previous one. Also the bit rate increasing allows to increase the encoding and decoding speed by 1.81-2.11 times. The developed system provides high reliability of data storage through the possibility of data recovery in case of some disks failure, and also high data security due to the distributed storage.
Keywords: residue number system, correction codes, storage, reliability.
References
- Garth A. Network attached storage architecture / Garth A. Gibson and Rodney Van Meter // Communications of the ACM. – New York, 2000. – Vol. 43 (№ 11). – P. 37–45.
- Guangyan Zhang. RAID+: Deterministic and Balanced Data Distribution for Large Disk Enclosures / Guangyan Zhang, Zican Huang, Xiaosong Ma, Songlin Yang, Zhufan Wang, Weimin Zheng // This paper is included in the Proceedings of the 16th USENIX Conference on File and Storage Technologies. – Oakland, February 12–15, 2018. – R. 280–283.
- Wan J. Thin RAID: Thinning Down RAID Array for Energy Conservation / J. Wan, X. Qu, N. Zhao, J. Wang and C. Xie // IEEE Transactions on Parallel & Distributed Systems. – 2015. – Vol. 26 (№ 10). – P. 2903–2915.
- Rahman P. A. Reliability model of disk arrays RAID-5 with data striping / P. A. Rahman and G. DK Novikova, Freyre Shavier // IOP Conf. Series: Materials Science and Engineering. – 2018. – Vol. 327. – R. 788–794.
- Keval Kachhala. RAID (Redundant Array of independent Disks) / Keval Kachhala, Rupali Gangarde // International Journal of Engineering Trends and Technology (IJETT). – May 2016. – Vol. 35 (№ 12). – R. 574–577.
- Yongkun Li. Analysis of Reliability Dynamics of SSD RAID / Yongkun Li, P. C. Lee, John C. S. Lui // IEEE TRANSACTIONS ON COMPUTERS. – 2014. – R. 1–14.
- Antonio Celesti. Adding long-term availability, obfuscation, and encryption to multi-cloud storage systems / Antonio Celesti, Maria Fazio, Massimo Villari, Antonio Puliafito // Journal of Network and Computer Applications. – January 2016. – Vol. 59. – P. 208–218.
- Tay Thian Fatt. A new algorithm for single residue digit error correction in Redundant Residue Number System / Tay Thian Fatt // Circuits and Systems (ISCAS), IEEE International Symposium IEEE. – 2014. – P. 1748–1751.
- Alsheikh M. A. Rate-distortion Balanced Data Compression for Wireless Sensor Networks / Alsheik M. A., Lin S., Niyato D., & Tan H. P. // IEEE Sensors Journal. – 2016. – Vol. 16 (№12). – P. 5072–5083.
- Xiao H. New Error Control Algorithms for Residue Number System Codes / Xiao H., Garg H. K., Hu J., & Xiao G // ETRI Journal. – 2016. – Vol. 38 (№ 2). – P. 326–336.
- Akushskij I. Ya. Mashinnaya arifmetika v ostatochnyh klassah / Akushskij I. Ya., Yudickij D.I. – M. : Sov. radio. 1968. – S. 460.
- Kulyna S.V. Vyiavlennia pomylok na osnovi koryhuiuchykh kodiv systemy zalyshkovykh klasiv / S.V. Kulyna // Fizyko-tekhnolohichni problemy peredavannia, obroblennia ta zberihannia informatsii v infokomunikatsiinykh systemakh : materialy VII Mizhnarodnoi naukovo-praktychnoi konferentsii. – Chernivtsi : «Misto», 2018. – S. 126–127.
