Workflow Resource Manager for Cloud Computing

Solution Overview

Problem

Conventional cloud computing environments face challenges in effectively provisioning resources for resource-intensive applications like AI and ML workloads, leading to workflow failures due to insufficient resources, increased costs, and reduced efficiency.

Innovation Solution

A workflow resource manager that analyzes resource requirements for each layer of a workflow and pre-provisions resources, ensuring availability before execution, thereby preventing failures and optimizing resource utilization across the cloud computing system.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If cloud computing environments provision resources dynamically during workflow execution, then resource flexibility is improved, but workflow failures occur due to insufficient resources available at execution time

Engineering Contradiction:

Improveresource flexibilityVSAvoidworkflow execution success

Core Design Contradiction:

Adaptability or versatilityVSReliability

Solution Approach 1:

The system performs preliminary resource provisioning by analyzing workflow definitions before execution and allocating required resources in advance. The workflow resource manager determines resource requirements for each operation and provisions them prior to workflow start, ensuring resources are available when needed while maintaining the ability to adapt to different workflow types.

Inventive Principle:

Principle #10Preliminary action

2Reliability

If cloud computing environments allocate sufficient resources for all workflow operations, then workflow execution reliability is improved, but resource costs increase due to over-provisioning and idle time

Engineering Contradiction:

Improveworkflow execution successVSAvoidresource cost

Core Design Contradiction:

ReliabilityVSLoss of energy

Solution Approach 1:

The system performs preliminary resource provisioning by analyzing workflow definitions before execution and allocating required resources in advance. The workflow resource manager determines resource requirements for each operation and provisions them prior to workflow start, ensuring resources are available when needed while maintaining the ability to adapt to different workflow types.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system dynamically adjusts resource allocation parameters based on actual workflow execution progress and resource utilization. The workflow resource manager monitors resource usage and modifies provisioning parameters in real-time, transitioning from static over-provisioning to dynamic adaptive allocation that matches actual needs.

Inventive Principle:

Principle #35Parameter changes

3Productivity

If cloud computing environments provision resources on-demand during workflow execution, then resource utilization efficiency is improved, but workflow execution time increases due to resource provisioning delays

Engineering Contradiction:

Improveresource utilization efficiencyVSAvoidworkflow execution time

Core Design Contradiction:

ProductivityVSLoss of time

Solution Approach 1:

The system performs preliminary resource provisioning by analyzing workflow definitions before execution and allocating required resources in advance. The workflow resource manager determines resource requirements for each operation and provisions them prior to workflow start, ensuring resources are available when needed while maintaining the ability to adapt to different workflow types.

Inventive Principle:

Principle #10Preliminary action

Data Source

PatentUS11934877B2Workflow resource management for cloud computing services

Publication Date: 2024.03.19 RED HAT LLC

AI summary

A workflow resource manager receives a request to execute a workflow in a cloud computing environment. The workflow resource manager determines that a first set of cloud computing resource requirements associated with the first set of operations for the workflow is satisfied by available cloud computing resources, and responsive to determining that a second set of cloud computing resource requirements associated with a subsequent set of operations for the workflow is not satisfied by the available cloud computing resources, rejects the request to execute the workflow.