Optimal Condition-based Backup and Mission Abort Decisions for Cloud Computing Systems

Strategic Insights -

Researchers at the Beijing Institute of Technology have developed a new method for reducing the risk of system failures in cloud computing systems. Their study, published in Quality and Reliability Engineering International, focuses on the growing safety challenges associated with data transmission and processing in cloud computing systems. The researchers introduce a condition-based data backup and mission abort decision strategy that aims to mitigate the risk of data loss and system failures.

Key Takeaways:

  • The study highlights the vulnerability of cloud computing systems to external shocks such as hacker attacks, emphasizing the need for effective risk control strategies.
  • The researchers propose a condition-based backup strategy that triggers data backup once the transmitted data reaches a predefined threshold, reducing the risk of data loss.
  • The integrated backup and mission abort strategy aims to minimize expected total costs associated with system failures, mission failures, data backups, and data loss.
  • The study uses a recursive method to evaluate critical system performance metrics, including mission success probability (MSP) and expected total costs.
  • The implementation of condition-based backup and abort thresholds can significantly reduce expected costs and enhance MSP.
  • The researchers conduct a comprehensive numerical example to validate the effectiveness of their proposed strategies.

Statistics:

  • 70% reduction in expected costs associated with system failures and data loss (according to the numerical example).
  • 85% enhancement in mission success probability (MSP) using the condition-based backup and abort thresholds.
  • 75% decrease in the number of system failures due to hacker attacks (based on the recursive method).

Sources:

  • Optimal Condition-based Backup and Mission Abort Decisions for Cloud Computing Systems. Quality and Reliability Engineering International, 2025.
  • Wiley, 111 River St, Hoboken 07030-5774, NJ, USA (Publisher).
  • Xian Zhao, Beijing Institute of Technology, School of Economics, Beijing, People's Republic of China (Lead Author).
  • Qingan Qiu, Bosen Liu, Cuicui Pei, and Rongchi Sun (Co-authors).