Task Scheduling Modes for Mixed Online and Offline Workloads
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current technologies lack an effective method to schedule online and offline tasks efficiently, especially in scenarios where they are mixedly deployed on hardware resources, leading to inefficiencies in resource utilization and delay requirements.
Innovation Solution
A method and apparatus for task scheduling that prioritizes online tasks, utilizing a sequential, throughput, or round-robin mode to ensure timely processing and adapt to various scenarios, with a hierarchical queue system dividing tasks into online and offline sets with varying priorities and resource requirements.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If online and offline tasks are mixedly deployed on hardware resources, then resource utilization is improved, but task processing delay and scheduling efficiency deteriorate
Solution Approach 1:
The patent divides tasks into two distinct sets: online task set and offline task set. Each set is scheduled independently with different scheduling modes, allowing online tasks to receive preferential treatment for timely processing while offline tasks utilize available resources without interfering with online task deadlines. This segmentation resolves the contradiction by preventing mixed deployment from causing delays.
Solution Approach 2:
The patent implements dynamic scheduling mode selection based on task characteristics. The scheduler can switch between sequential mode, throughput mode, and round-robin mode depending on the task set and current system state. This dynamic adaptation allows optimal scheduling that balances resource utilization with timely processing requirements.
2Ease of operation
If a single scheduling mode is used for all tasks, then scheduling simplicity is maintained, but adaptability to different scenarios deteriorates
Solution Approach 1:
The patent implements dynamic scheduling mode selection based on task characteristics. The scheduler can switch between sequential mode, throughput mode, and round-robin mode depending on the task set and current system state. This dynamic adaptation allows optimal scheduling that balances resource utilization with timely processing requirements.
Solution Approach 2:
The patent changes the scheduling parameter (mode) based on task type and system state. By adjusting the scheduling mode parameter dynamically, the system adapts to different scenarios while maintaining a relatively simple overall scheduling framework. The scheduler monitors task characteristics and selects appropriate modes without requiring complex manual configuration.
Data Source
AI summary
Embodiments of the present disclosure relate to a method for task scheduling and an electronic device. The method comprises: based on determining there is a task to be scheduled in an online task set of a waiting queue, an electronic device schedules the task based on a preset online scheduling mode which may be a sequential mode, a throughput mode, or a round-robin mode. Accordingly, at least one processing engine of the electronic device can process the task.


