Primary VM Fault Selection for Stable Fault-Tolerant Switching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional fault-tolerant systems fail to differentiate between hardware and software faults, leading to unnecessary control switching and potential system instability when software faults are present.
Innovation Solution
A primary machine equipped with a synchronization information generator and fault selector to determine the type of fault, allowing controlled switching only when necessary, such as in hardware faults, while software faults are handled locally.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If control switching is performed for all faults, then system reliability is improved, but unnecessary switching occurs for software faults causing system instability
Solution Approach 1:
The fault handling mechanism is segmented into two distinct paths: one for hardware faults that triggers control switching, and another for software faults that handles errors locally. The fault selector divides fault information processing based on fault type, preventing unnecessary control switching for software faults while maintaining reliability for hardware faults.
2Reliability
If control switching is performed for all faults, then fault tolerance is improved, but operation continuity deteriorates due to unnecessary switching
Solution Approach 1:
The system dynamically adjusts its fault handling strategy based on the type of fault detected. For hardware faults, the system activates control switching to maintain fault tolerance. For software faults, the system activates local error handling to preserve operation continuity. This dynamic adaptation allows the system to optimize between fault tolerance and operation continuity based on real-time conditions.
3Productivity
If fault differentiation is implemented, then unnecessary switching is reduced, but device complexity increases
Solution Approach 1:
The fault information processing is segmented into distinct handling paths based on fault type. The fault selector acts as a decision node that divides fault information into hardware fault categories and software fault categories, directing each to appropriate handling mechanisms. This segmentation reduces unnecessary switching operations while maintaining manageable system complexity through structured organization.
4Reliability
If control switching is performed for all faults, then system robustness is improved, but load increases due to frequent switching
Solution Approach 1:
Instead of applying control switching to all faults (excessive action), the system applies control switching only to hardware faults where it is necessary (partial action). For software faults, local error handling suffices, avoiding the additional load of control switching. This partial application of the fault tolerance mechanism maintains system robustness for critical hardware failures while reducing overall system load.
Data Source
AI summary
A primary machine includes a primary virtual machine including a synchronization information generator configured to generate and output synchronization information based on an instruction and a result of execution of the instruction, and a fault selector configured to determine a type of fault information generated when the instruction was executed. The primary VM changes operation depending on a result of the determination of the type of fault information.