Virtual Machine Scheduling via Constraint-Based Migration Paths
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing virtual machine scheduling algorithms are strongly coupled with scheduling requirements, leading to increased costs and reduced flexibility when changes occur, necessitating adjustments to the underlying algorithm.
Innovation Solution
A method that decouples the virtual machine scheduling algorithm from service-side requirements by converting scheduling needs into constraint conditions using a pre-configured conversion rule, allowing the same algorithm to adapt to various scheduling objectives through affinity coefficients and other constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Manufacturing precision
If a scheduling algorithm is strongly coupled with virtual machine scheduling requirements, then the algorithm can precisely meet specific scheduling objectives, but the algorithm complexity and adjustment costs increase when requirements change
Solution Approach 1:
The patent segments the scheduling system into two independent parts: a requirement conversion module that translates service-side scheduling requirements into constraint conditions, and a basic scheduling algorithm module that handles the actual scheduling. This segmentation allows the algorithm to remain simple and stable while the conversion module adapts to different requirements, resolving the contradiction between precision and complexity.
Solution Approach 2:
The patent introduces constraint conditions as an intermediary layer between scheduling requirements and the scheduling algorithm. The conversion rule transforms diverse service-side requirements into standardized constraint conditions that the basic algorithm can process uniformly. This intermediary enables precise requirement fulfillment without increasing algorithm complexity, as the algorithm only needs to handle standard constraints rather than diverse requirements directly.
2Adaptability or versatility
If the scheduling algorithm is adjusted to meet changing virtual machine scheduling requirements, then the scheduling can adapt to new objectives, but the adjustment costs and time increase
Solution Approach 1:
The patent performs preliminary action by pre-defining conversion rules that map various service-side scheduling requirements to standardized constraint conditions. When new scheduling requirements arise, the system only needs to apply the appropriate pre-defined conversion rule rather than developing or adjusting the entire scheduling algorithm. This preliminary preparation enables rapid adaptation to changing requirements without time-consuming algorithm adjustments.
Solution Approach 2:
The basic scheduling algorithm is designed with universality to handle multiple types of scheduling requirements through a unified interface. By converting diverse requirements into standardized constraint conditions, the same basic algorithm can serve multiple scheduling objectives (such as load balancing, energy efficiency, service level agreement compliance) without needing separate algorithms or adjustments, thereby achieving high adaptability with minimal adjustment time.
3Adaptability or versatility
If multiple scheduling algorithms are developed for different service requirements, then each service can have optimized scheduling, but the system complexity and maintenance costs increase
Solution Approach 1:
The patent implements a universal scheduling system where a single basic scheduling algorithm serves multiple service requirements. The conversion rule acts as an adapter that translates service-specific requirements into the standardized constraint format the basic algorithm expects. This approach achieves comprehensive service coverage while maintaining low system complexity, as there is only one algorithm to implement, test, and maintain rather than multiple service-specific algorithms.
Solution Approach 2:
Instead of creating multiple distinct algorithms for different services, the patent uses the copying approach by replicating the conversion rule layer for different service requirements while reusing the same basic scheduling algorithm. Each service requirement has its own conversion rule set that copies the pattern of translating requirements into constraints, but the core algorithm remains a single shared component, reducing overall system complexity while maintaining service-specific optimization.
Data Source
AI summary
The present disclosure relates to a virtual machine scheduling method, a medium, and an electronic device, and relates to the field of computer technologies. The virtual machine scheduling requirement is converted into a target constraint condition through resource information of a target virtual machine and a target physical machine, and a target virtual machine migration path is determined according to the target constraint condition, the resource information of the target virtual machine and the target physical machine, and a basic scheduling algorithm. The virtual machine scheduling is performed according to the target virtual machine migration path.