Task Scheduling Token Mechanism for Computing Resource Utilization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional task scheduling methods in computing systems face performance deterioration due to blocking and contention issues, especially when operations like sleep, mutex, lock, and read/write occur, leading to inefficient utilization of computing resources.
Innovation Solution
A novel task scheduling method that utilizes a token mechanism to manage computing resources, allowing for accurate tracking of resource states and efficient task execution by switching between tasks based on priority and availability of tokens, thereby optimizing resource utilization without relying on system kernel customization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If kernel customization is used to implement task scheduling, then computing resource utilization is maximized, but system complexity and customization requirements increase
Solution Approach 1:
The patent introduces a task scheduling token as an intermediary mechanism that mediates between multiple computing resources and tasks. The token acts as a carrier of scheduling authority, allowing resources to acquire and release tokens without complex kernel customization. This intermediary simplifies the scheduling mechanism while maintaining efficient resource utilization.
Solution Approach 2:
The patent creates multiple copies of task scheduling tokens that can be distributed among computing resources. Instead of a single centralized scheduler, each resource can hold a token copy, enabling decentralized task execution. This copying approach reduces system complexity by eliminating the need for centralized kernel management while maintaining productivity.
2Reliability
If computing resource executes blocking operations (sleep, mutex, lock, read/write), then task completion may be achieved, but other waiting tasks experience increased wait time and system performance deteriorates
Solution Approach 1:
The patent implements preliminary action by pre-creating a pool of task scheduling tokens before task execution begins. When a computing resource becomes blocked, other resources can immediately acquire available tokens from the pool to execute waiting tasks, without needing to wait for the blocked resource to become available. This eliminates the time loss caused by blocking operations.
Solution Approach 2:
The patent ensures continuity of useful action by maintaining a pool of idle tokens that can be immediately assigned to waiting tasks. When one computing resource is blocked performing useful work, other resources with idle tokens can continue executing tasks, ensuring that the system's productive activity continues without interruption or idle time.
3Productivity
If a computing resource acquires a task scheduling token to execute a task, then task execution is enabled, but the resource cannot switch to higher priority tasks without discarding and retrieving tokens
Solution Approach 1:
The patent implements dynamics by making token assignment flexible and adaptive. Computing resources can dynamically acquire and release tokens based on task priority requirements. When a higher priority task becomes available, a resource can discard its current token, execute the priority task, and then retrieve a token to continue the original task. This dynamic token management enables priority-based scheduling while maintaining execution efficiency.
Data Source
AI summary
A method for task scheduling includes, if it is determined that there is an idle task scheduling token, causing a first computing resource in a computing system to acquire the idle task scheduling token as a first task scheduling token to execute a first task in a task queue. If it is determined that a second task having a higher priority than the first task is to be executed, the first computing resource is to discard the first task scheduling token to switch from executing the first task to executing the second task. If it is determined that the execution of the second task is finished, the first computing resource is to retrieve the first task scheduling token to continue the execution of the first task.