Dynamic Resource Provisioning for Processing Nodes

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing resource provisioning technologies face challenges in accurately determining task demands, leading to resource usage imbalances and high power consumption due to static resource allocation methods that do not account for dynamic task characteristics.

Innovation Solution

A method and apparatus for dynamic resource provisioning that involves obtaining component metric information of processing nodes and task characteristics, allowing for the selection and deployment of the most suitable nodes based on real-time resource requirements, thereby optimizing resource usage and reducing power consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If static resource allocation methods are used, then device complexity is reduced, but resource usage balance deteriorates and power consumption increases

Engineering Contradiction:
Improveresource allocation methodVSAvoidpower consumption
Core Design Contradiction:
Device complexityVSLoss of energy

Solution Approach 1:

The patent implements dynamic resource provisioning by continuously monitoring task characteristics and component metric information, allowing the system to adapt resource allocation in real-time based on actual task demands rather than using fixed static allocation methods

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system establishes a feedback mechanism where task characteristics are obtained during execution, compared with component metric information, and used to determine optimal processing nodes for resource allocation, creating a closed-loop control system that reduces energy waste

Inventive Principle:
Principle #23Feedback

2Device complexity

If static resource allocation methods are used, then device complexity is reduced, but resource usage rate deteriorates

Engineering Contradiction:
Improveresource allocation methodVSAvoidresource usage rate
Core Design Contradiction:
Device complexityVSProductivity

Solution Approach 1:

The system dynamically adjusts resource allocation based on real-time task characteristics and component metric information, enabling optimal matching between task demands and available resources to maximize resource usage rate

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The patent changes the allocation parameters from static fixed values to dynamic values that are continuously adjusted based on task characteristics such as CPU usage, memory requirements, and I/O patterns, improving resource utilization efficiency

Inventive Principle:
Principle #35Parameter changes

3Measurement precision

If task characteristics are not accurately determined, then measurement precision is reduced, but resource provisioning simplicity is improved

Engineering Contradiction:
Improvetask demand determination accuracyVSAvoidresource provisioning system
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The system performs preliminary measurement of task characteristics during the execution phase on a first set of processing nodes, gathering data about actual resource consumption patterns before making final provisioning decisions

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary measurement and analysis layer that collects task characteristics and component metric information, processes this data to determine optimal resource allocation, and then provisions resources accordingly, separating the complexity of accurate measurement from the provisioning decision-making process

Inventive Principle:
Principle #24Intermediary (Mediator)

Data Source

PatentUS10789100B2System, apparatus and method for resource provisioning
Publication Date: 2020.09.29 CLOUD INTELLIGENCE ASSETS HOLDING (SINGAPORE) PTE LTD
  • US10789100B2 patent drawing
  • US10789100B2 patent drawing
  • US10789100B2 patent drawing

AI summary

A method of resource provisioning including obtaining component metric information of one or more processing nodes, where the one or more processing nodes form a pool of processing nodes managed by the provisioning apparatus. The method also includes obtaining task characteristics of a target task executing on one or more processing nodes of a first set, where the one or more processing nodes of the first set are selected from the pool of processing nodes. The method further includes determining one or more processing nodes of a second set from the pool of processing nodes based on the task characteristics and the component metric information and the step of deploying the target task to the one or more processing nodes in the second set.