Hierarchical Edge Workload Balancing via Dual Schedulers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing edge computing systems face inefficiencies in workload management, leading to idle resources and increased latency due to the lack of effective balancing of workloads among client devices and cloud computing resources, resulting in higher costs and potential stability issues.
Innovation Solution
A method involving the determination of workload criteria, division of workloads into groups, assignment of edge computing resources, and the use of first and second-level schedulers to balance and manage workloads dynamically, ensuring efficient distribution and handling of workloads across edge computing resources, with error detection and mitigation mechanisms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If workloads are not balanced among edge computing resources, then resource utilization is low and idle resources increase, but implementing workload balancing increases system complexity and requires multiple schedulers
Solution Approach 1:
The system divides workload management into two hierarchical levels: a first scheduler at the control center that divides workloads into groups and assigns them to edge computing resources, and second schedulers at each edge resource that further divide and balance workloads locally. This segmentation allows each scheduler to operate independently within its scope, improving resource utilization without requiring a single complex centralized system.
Solution Approach 2:
The patent introduces a hierarchical dimension to workload scheduling by adding a second level of schedulers at the edge computing resources. This transforms the traditional single-level centralized scheduling into a multi-level distributed system, enabling workload balancing to occur both centrally and locally, thereby improving resource utilization while distributing system complexity across multiple manageable components.
2Loss of time
If workloads are concentrated in fewer edge computing resources, then latency is reduced, but system reliability decreases and stability issues arise
Solution Approach 1:
The system segments workloads into multiple groups and distributes them across different edge computing resources through hierarchical scheduling. This prevents workload concentration on single resources, reducing the risk of system failures while maintaining low latency through optimized local scheduling at each edge resource.
Solution Approach 2:
The patent dynamically adjusts workload distribution parameters based on system conditions, utilizing load balancing algorithms that modify allocation strategies in real-time. This allows the system to optimize latency performance while maintaining reliability through adaptive parameter changes rather than fixed concentration patterns.
3Productivity
If more intermediary servers are used to manage workloads, then workload distribution improves, but system complexity and costs increase
Solution Approach 1:
The second schedulers implemented at each edge computing resource perform multiple functions: they receive workload groups from the first scheduler, divide these workloads further, balance loads locally across available resources, and manage error detection and mitigation. This multi-functionality eliminates the need for separate specialized intermediary servers, improving workload distribution efficiency while reducing system complexity and costs.
Solution Approach 2:
Each edge computing resource is equipped with its own second scheduler that autonomously manages workload division and balancing without requiring constant centralized control. This self-service capability allows efficient workload distribution at the edge while reducing the need for additional intermediary servers, thereby lowering system complexity and operational costs.
4Device complexity
If centralized workload management is used, then control is simplified, but latency increases and resource responsiveness decreases
Solution Approach 1:
The system segments workload management authority between a first scheduler at the control center for high-level workload grouping and assignment, and second schedulers at each edge resource for local workload division and balancing. This segmentation maintains simplified centralized control for strategic decisions while enabling fast local responses for operational adjustments, thereby reducing latency without complicating overall management control.
Solution Approach 2:
The patent adds a hierarchical dimension to workload management, creating a two-level control structure that combines centralized strategic control with decentralized tactical execution. This hierarchical approach maintains the simplicity of centralized management for workload group assignment while enabling low-latency local responses through edge-based schedulers, effectively resolving the contradiction between control simplicity and response speed.
Data Source
AI summary
A set of workload criteria is determined from a workload associated with a plurality of sources. The workload is divided among a set of workload groups according to the set of workload criteria and a first workload scheduler. A set of edge computing resources is assigned to each workload group within the set according to the set of workload criteria and the set of workload groups. A portion of the workload associated with a subset of the plurality of sources is handled by a first subset of edge computing resources and a second workload scheduler, where the subset of sources is associated with a first workload group. The handling includes balancing, by the second workload scheduler, the portion of the workload among the subset of sources. The handled workload is reported to a control center.