VM Scheduler Fixed Sequence Cycle Period
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing control units with virtualization systems face inefficiencies in scheduling and resource allocation, particularly in low-footprint, event-intensive, and real-time applications, due to high task switches, short execution times, and complex multi-core architectures, leading to hypervisor scheduler timeouts and increased overhead.
Innovation Solution
A control unit with a virtualization system that includes a scheduler assigning processing time to virtual machines in a predetermined fixed sequence based on real-time attributes, using a cycle period for repeated scheduling, allowing for different real-time attributes among virtual machines and optimizing task periods to reduce virtual machine switches and jitter.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If all virtual machines are scheduled one after another within a scheduling cycle, then each virtual machine receives processing time, but hypervisor scheduler timeouts increase and scheduling overhead becomes high
Solution Approach 1:
The scheduling cycle is segmented into multiple slots, with each slot dedicated to scheduling a specific virtual machine. This segmentation allows the hypervisor to schedule virtual machines in an optimized sequence rather than uniformly one after another, reducing the number of scheduling operations and timeouts while ensuring each VM receives its required processing time.
2Speed
If the number of task switches is increased to handle high event intensity, then system responsiveness improves, but processor performance decreases due to excessive switching overhead
Solution Approach 1:
The system uses periodic scheduling slots to manage virtual machine execution. By organizing task execution into periodic time slots assigned to different virtual machines, the system maintains high responsiveness for time-critical tasks while reducing the frequency and overhead of context switches compared to event-driven switching.
3Power
If multi-core architecture is used to increase computing power, then processing capacity improves, but system complexity increases making resource distribution difficult
Solution Approach 1:
The patent introduces a virtualization layer with a hypervisor that acts as an intermediary between the physical multi-core processor and virtual machines. This intermediary manages the complex task of distributing tasks across multiple cores, handling core affinity, and balancing load, thereby enabling multi-core utilization without exposing the full scheduling complexity to individual virtual machines or applications.
Data Source
AI summary
A control unit including at least one processor and at least one memory connected to the at least one processor, a virtualization system, the virtualization system including a scheduler for scheduling a plurality of virtual machines to assign processing time to each of the virtual machines according to a predetermined fixed sequence of virtual machine switches forming a cycle period, which is repeated, the cycle period being the minimum time period after which the scheduling is repeated, wherein the virtualization system and the plurality of virtual machines are real-time systems, the virtual machines having respectively at least one real-time attribute, wherein at least one the real time attribute of a first virtual machine are different to the corresponding real-time attribute(s) of a second virtual machine, wherein predetermined fixed sequence of virtual machine switches is calculated based on the at least one real time attribute.