MODELING AND ASSESSMENT OF RELIABLE CLOUD STORAGE SOLUTIONS FOR CULTURAL HERITAGE ASSETS
Keywords:
Reliability, MARS, Azure, disaster recovery, cultural heritageAbstract
This paper presents a comprehensive framework for evaluating the performance and reliability of a fully cloud hosted disaster recovery solution dedicated to the long term preservation of cultural heritage data. The study focuses on Microsoft Azure Recovery Service MARS, a platform that provides continuous, encrypted backup and granular recovery options designed to protect critical information against natural disasters, cyberattacks, and hardware failures. To ensure realistic results, we conducted year long empirical measurements in a live production environment that mirrors the operational conditions of institutions responsible for safeguarding digital cultural assets. Backup and restore jobs were executed on a dedicated virtual machine connected through a symmetrical 200/200 Mbps Internet link, allowing the collection of detailed metrics such as average data throughput, Recovery Point Objective RPO, Recovery Time Objective RTO, and monthly service cost. These measurements capture the true performance envelope of the cloud service, including the impact of network fluctuations and encryption overhead.
Alongside performance monitoring, a system dynamics reliability model was developed to quantify end to end service resilience.
The model represents the on premises data center, the Internet service provider ISP, and the Azure cloud as a series configuration and applies classical reliability theory to calculate time dependent availability. Mean time between failures and published service level agreements SLA were used to parameterize the simulation.
Results reveal that while Azure infrastructure consistently delivers near perfect service availability (approaching eleven nines of uptime) the overall system reliability is ultimately governed by the weakest component of the delivery chain.
Over a 15 day observation window the integrated system achieved an aggregate reliability of approximately 0.97, with the ISP connection identified as the dominant limiting factor. These findings provide a practical foundation for designing resilient digital repositories for cultural heritage institutions. They emphasize the need for redundant high
bandwidth connectivity and continuous validation of recovery procedures to maintain strict RPO and RTO objectives. When such network safeguards are combined with the security, encryption, and long term retention features of Microsoft Azure Recovery Service, organizations can confidently implement a sustainable, standards
compliant archival strategy capable of protecting irreplaceable cultural and historical artifacts for decades to come.
References
Joshi, S. C., & Sivalingam, K. M. (2014). Fault tolerance mechanisms for virtual data center architectures. Springer Science+Business Media, New York.
Liu, Z., et al. (2022). Reliability modelling and optimization for microservice-based cloud application using multi-agent system. IET Communications, 16(10), 1182–1199.
Mathew, A., & Mai, C. (2018). Study of various data recovery and data backup techniques in cloud computing & their comparison. Proceedings of the 3rd IEEE International Conference on Recent Trends in Electronics, Information & Communication Technology, Bangalore, India.
Mendonca, J., Lima, R., Andrade, E., Araujo, J., & Kim, D. S. (2020). Multiple-criteria evaluation of disaster recovery strategies based on stochastic models. 16th International Conference on the Design of Reliable Communication Networks (DRCN), IEEE.
Mendonca, J., Lima, R., Queiroz, E., Andrade, E., & Kim, D. S. (2019). Evaluation of a backup-as-a-service environment for disaster recovery. IEEE Symposium on Computers and Communications (ISCC), Barcelona, Spain.
Miller, H. E., & Engemann, K. J. (2014). Using reliability and simulation models in business continuity planning. International Journal of Technology, Policy and Management, 5(1), 43–56.
Nikolovski, S., Milenkovski, B., & Petreska, A. (2025). Reliability analysis through system dynamics modeling for intellectual capital protection with a cloud services. 60th International Scientific Conference on Information, Communication and Energy Systems and Technologies (ICEST), IEEE.
Nikolovski, S., Mitrevski, P., & Petreska, A. (2024). Performance analysis of cloud service-based data protection systems. 59th International Scientific Conference on Information, Communication and Energy Systems and Technologies (ICEST), IEEE.
Nikolovski, S., & Mitrevski, P. (2023). Data protection and recovery performance analysis of cloud-based recovery service. 58th International Scientific Conference on Information, Communication and Energy Systems and Technologies (ICEST), IEEE.
Rajini, N., et al. (2020). Reliability of cloud services provided to non-banking financial institutions. International Journal of Control and Automation, 13(2s), 165–172.
Rebah, H. B., & Sta, H. B. (2016). Disaster recovery as a service: A disaster recovery plan in the cloud for SMEs. Global Summit on Computer & Information Technology, Sousse, Tunisia.
Shahzadi, S., Ubakanma, G., Iqbal, M., & Dagiuklas, T. (2018). Autonomous, seamless and resilience carrier cloud brokerage solution for business contingencies during disaster recovery. IEEE 20th International Conference on High Performance Computing and Communications, Exeter, UK.
Tamimi, A. A., Dawood, R., & Sadaqa, L. (2019). Disaster recovery techniques in cloud computing. IEEE Jordan International Joint Conference on Electrical Engineering and Information Technology (JEEIT), Amman, Jordan.
Tian, Y., Tian, J., & Li, N. (2020). Cloud reliability and efficiency improvement via failure risk based proactive actions. Journal of Systems and Software, 163, 110524.
Wood, T., Cecchet, E., Ramakrishnan, K., Shenoy, P., Van Der Merwe, J., & Venkataramani, A. (2010). Disaster recovery as a cloud service: Economic benefits & deployment challenges. 2nd USENIX Workshop on Hot Topics in Cloud Computing, Boston, MA.
