Decentralized Worker Process Traffic Control for Resource Overload

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

In cloud environments, the large number of worker processes (WPs) can overwhelm target resources, leading to excessive traffic and errors, which existing master schedulers attempt to manage but add complexity and consume resources.

Innovation Solution

Each WP autonomously determines, based on historical warnings from the target resource, whether accessing the resource would exceed a dynamic threshold, and adjusts its actions accordingly to prevent congestion without a centralized scheduler.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If a large number of worker processes are used to handle data lifecycle management tasks, then productivity is improved, but the target resource becomes overwhelmed by excessive traffic

Engineering Contradiction:

Improvedata lifecycle management task completion rateVSAvoidresource overload and traffic congestion

Core Design Contradiction:

ProductivityVSObject-affected harmful factors

Solution Approach 1:

Each worker process autonomously monitors historical warning indications from the target resource and uses this feedback to dynamically adjust its access behavior. When warnings indicate high traffic conditions, worker processes autonomously back off and reduce their access rate, creating a self-regulating system that prevents resource overload while maintaining high productivity.

Inventive Principle:

Principle #23Feedback

2Object-affected harmful factors

If a master scheduler is used to control the number of worker processes accessing the target resource, then resource overload is prevented, but device complexity and resource consumption increase

Engineering Contradiction:

Improveresource overload preventionVSAvoidsystem architecture complexity

Core Design Contradiction:

Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

Each worker process is equipped with autonomous decision-making capability to independently determine whether to access the target resource based on historical warning indications. This self-service approach eliminates the need for a centralized master scheduler, reducing system complexity and resource consumption while effectively preventing resource overload through distributed intelligence.

Inventive Principle:

Principle #25Self-service

3Device complexity

If worker processes autonomously access the target resource without coordination, then device complexity is reduced, but resource overload and errors increase

Engineering Contradiction:

Improvecoordination system complexityVSAvoiderror rate in data lifecycle management operations

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

Worker processes continuously monitor historical warning indications from the target resource and use this feedback to autonomously adjust their access behavior. This distributed feedback mechanism enables coordination without centralized control, maintaining low system complexity while significantly improving reliability by preventing resource overload and reducing error rates.

Inventive Principle:

Principle #23Feedback

Data Source

PatentUS12498969B2Distributed, decentralized traffic control for worker processes in limited-coordination environments

Publication Date: 2025.12.16 MICROSOFT TECHNOLOGY LICENSING LLC

AI summary

Solutions for distributed, decentralized traffic control for worker processes (WPs) in limited-coordination environments include: by each WP of a plurality of WPs: receiving, by the WP, indications of historical warnings corresponding to a target resource; based on at least the indications of historical warnings, autonomously determining, by the WP, whether the WP accessing the target resource would exceed a dynamic threshold of WPs permitted to access the target resource; based on at least determining that the WP accessing the target resource would exceed the dynamic threshold, not accessing the target resource by the WP; and based on at least determining that the WP accessing the target resource would not exceed the dynamic threshold, accessing the target resource, by the WP, to perform a first data management task.