Dynamic Task Propagation in Distributed Workload Agents

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed computing environments, workload agents often face capacity overload, leading to job delays or errors, as existing scheduling systems lack efficient mechanisms to dynamically redistribute tasks when an agent's resources are exceeded.

Innovation Solution

A method where workload agents can forward jobs to secondary agents based on workload parameters like CPU, memory, and throughput, using a forwarding map to optimize resource utilization and scale capacity without disrupting the scheduler.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If jobs are assigned to workload agents based on traditional scheduling, then the scheduler can manage task distribution, but workload agents experience capacity overload leading to job delays or errors

Engineering Contradiction:
Improvejob completion rateVSAvoidjob execution success rate
Core Design Contradiction:
ProductivityVSReliability

Solution Approach 1:

The patent implements dynamic task propagation where workload agents can forward jobs to secondary agents based on real-time capacity conditions. The forwarding mechanism allows the system to adapt dynamically to changing workload conditions, preventing capacity overload from causing job failures while maintaining high productivity through flexible task redistribution.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent introduces a forwarding mechanism that acts as an intermediary between the scheduler and workload agents. When primary agents cannot handle jobs, the forwarding map redirects tasks through intermediate agents to secondary agents with available capacity, ensuring reliable job execution without compromising scheduler management efficiency.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Device complexity

If the scheduler directly manages all task distribution, then centralized control is maintained, but scheduler overhead increases and scalability is limited

Engineering Contradiction:
Improvescheduler overheadVSAvoiddynamic task redistribution capability
Core Design Contradiction:
Device complexityVSAdaptability or versatility

Solution Approach 1:

The patent segments the centralized scheduling function by enabling workload agents to independently make forwarding decisions based on local capacity conditions. This segmentation reduces scheduler overhead by distributing intelligence to agents while maintaining adaptability through the forwarding map mechanism that enables dynamic task redistribution without centralized micromanagement.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements self-service capability in workload agents through the forwarding mechanism. Agents autonomously evaluate their capacity conditions and forward jobs to secondary agents when necessary, eliminating the need for constant scheduler intervention and reducing overhead while maintaining high adaptability to changing workload conditions.

Inventive Principle:
Principle #25Self-service

3Productivity

If workload capacity is fixed for each agent, then resource allocation is simple, but the system cannot optimize resource utilization under varying workload conditions

Engineering Contradiction:
Improveresource utilization efficiencyVSAvoidcapacity management complexity
Core Design Contradiction:
ProductivityVSDevice complexity

Solution Approach 1:

The patent transforms fixed capacity management into a dynamic system where agents can leverage secondary agents' resources when capacity is insufficient. The forwarding map enables real-time capacity optimization by routing jobs to agents with available resources, improving resource utilization efficiency without requiring complex centralized capacity management.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent makes workload agents multi-functional by enabling them to serve as both primary agents for their own jobs and secondary agents for forwarding jobs from other agents. This universality improves resource utilization efficiency as any agent can contribute its spare capacity to the system while keeping capacity management relatively simple through the standardized forwarding mechanism.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Data Source

PatentUS10146584B2Weight adjusted dynamic task propagation
Publication Date: 2018.12.04 CA TECH INC
  • US10146584B2 patent drawing
  • US10146584B2 patent drawing
  • US10146584B2 patent drawing

AI summary

A method of assigning data processing tasks to workload agents includes receiving workload parameters that relate to available capacities of a plurality of workload agents, identifying a primary workload agent based on at least one of the workload parameters; identifying a plurality of candidate secondary workload agents; identifying a secondary workload agent based on the plurality of workload parameters; and transmitting a job message that contains a command for the primary workload agent to perform a data processing task. The job message includes a forwarding map that identifies the secondary workload agent, and contains a command for the primary workload agent to perform the data processing task using resources of the secondary workload agent.