一种基于分层抽象模型的任务调度方法及系统
By generating dual-track error code rules, action dependency rules, and exponential backoff retry rules, combined with stateless atomic actions and breakpoint snapshot metadata, the lock contention and architecture compatibility issues in the task scheduling architecture are resolved. This achieves efficient cross-node state coordination and breakpoint resume, ensuring execution performance and process reliability in large-scale parallel scenarios.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- CHINA UNICOM INTERNET OF THINGS CO LTD
- Filing Date
- 2026-05-21
- Publication Date
- 2026-07-17
AI Technical Summary
Existing task scheduling architectures suffer from lock contention, deadlock, and thundering herd effects in large-scale node parallel scenarios. Furthermore, existing solutions may violate the layered decoupling principle of orchestration architecture or introduce additional external dependencies, leading to performance degradation and architecture compatibility issues.
A task scheduling method based on a hierarchical abstract model is adopted to generate dual-track error code rules, action dependency rules, and exponential backoff retry rules. Through stateless atomic actions and breakpoint snapshot metadata fields, cross-node state collaboration and breakpoint resume compatibility are achieved, avoiding modification of the original architecture logic.
Without intruding on the core logic of the original architecture, cross-node state collaboration and breakpoint resume compatibility were achieved, reducing system complexity and ensuring execution performance and process reliability in large-scale parallel scenarios.
Smart Images
Figure CN122220079B_ABST