Simple Hypervisor Resource Scheduling for Guest OS Continuity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In hypervisor-based resource scheduling, when a fault occurs, all guest OSs cannot continue operating due to the failure of resource dispatching and preempting processes, leading to a loss of access to essential CPU and memory resources.
Innovation Solution
A simple hypervisor is introduced alongside the primary hypervisor, which includes an agent function to allocate or deallocate resources to guest OSs, ensuring continued operation of critical guest OSs by emulating physical resources and performing resource scheduling, even if the primary hypervisor fails.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If a hypervisor is used to manage multiple guest OSs with dynamic resource scheduling, then resource allocation efficiency is improved, but system reliability deteriorates because a hypervisor fault causes all guest OSs to fail
Solution Approach 1:
The system segments the hypervisor functionality by introducing a simple hypervisor alongside the original hypervisor. The simple hypervisor handles only essential resource allocation tasks for critical guest OSs, while the original hypervisor manages non-critical workloads. This segmentation ensures that a fault in the original hypervisor does not affect the operation of critical guest OSs.
Solution Approach 2:
The invention changes the complexity parameter of the hypervisor by creating a simplified version with reduced functionality. The simple hypervisor implements only the necessary resource scheduling functions for high-availability guest OSs, excluding advanced features that increase complexity and potential failure points. This parameter change maintains essential resource allocation while improving reliability.
2Reliability
If a simple hypervisor is introduced to ensure high availability, then system reliability is improved, but device complexity increases
Solution Approach 1:
The invention extracts only the essential resource allocation functions from the complete hypervisor suite and implements them in a separate simple hypervisor. By taking out only the necessary scheduling and resource management capabilities required for high availability, the system achieves improved reliability without adding excessive complexity. The simple hypervisor contains minimal code and functionality, reducing the attack surface and potential failure modes.
Data Source
AI summary
A simple hypervisor, in addition to a hypervisor, is operated on a computer. A guest OS, the continued operations of which need to be guaranteed, when a fault occurs in the hypervisor is operated on the simple hypervisor, and the other guest OSs are operated on the hypervisor. The hypervisor performs resource scheduling (determining of resources to be allocated to or deallocated from each guest OS) and the simple hypervisor executes, in place of the simple hypervisor, allocation or deallocation of resources to or from the guest OS, the continued operations of which need to be guaranteed.