Data Provider Director for Grid Computing Cache Optimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Grid computing systems face high network overhead in providing data to remote engines, leading to inefficient data retrieval and increased bandwidth requirements.

Innovation Solution

Implementing a data provider system architecture with caching mechanisms, where data is stored in internal and external caches, and a data provider director determines the optimal data provider, minimizing network calls by utilizing caches and alternate providers to reduce data retrieval from sources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If data is retrieved from remote data sources for each computation, then data availability is ensured, but network bandwidth consumption increases and retrieval efficiency decreases

Engineering Contradiction:
Improvedata retrieval efficiencyVSAvoidnetwork bandwidth consumption
Core Design Contradiction:
ProductivityVSLoss of energy

Solution Approach 1:

The patent implements caching mechanisms that pre-fetch and store data from remote sources before it is needed for computations. Data providers cache data locally, and data provider directors maintain directories of available data in advance, eliminating the need for repeated network retrievals and reducing bandwidth consumption while improving retrieval efficiency

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces data provider directors as intermediary components between engines and data sources. These directors maintain local directories of data locations and coordinate data provision, reducing direct network traffic between engines and remote data sources while improving data availability and retrieval efficiency

Inventive Principle:
Principle #24Intermediary (Mediator)

2Loss of energy

If data is cached locally at data providers, then network bandwidth is reduced, but system complexity increases due to cache management

Engineering Contradiction:
Improvenetwork bandwidth consumptionVSAvoidcache management complexity
Core Design Contradiction:
Loss of energyVSDevice complexity

Solution Approach 1:

The patent divides the caching function into multiple components: local caches at individual data providers, external caches at data provider directors, and directory services. This segmentation distributes cache management responsibilities, reducing the complexity burden on any single component while achieving overall bandwidth reduction

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The data provider director serves multiple functions: it acts as a cache for frequently accessed data, maintains a directory of data locations across the grid, coordinates data provision requests, and manages lease agreements. This multi-functionality consolidates complexity into a single component rather than distributing it across multiple data providers

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

3Reliability

If multiple data providers are used with external caches, then data availability is improved, but coordination overhead increases

Engineering Contradiction:
Improvedata availabilityVSAvoidcoordination overhead
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent implements lease-based feedback mechanisms where data providers and directors negotiate and track data provision agreements. When data is cached externally, lease agreements track availability and expiration, providing automated feedback that coordinates multiple data providers without requiring complex manual management, thereby improving reliability while controlling coordination overhead

Inventive Principle:
Principle #23Feedback

Data Source

PatentUS7882200B2Method and apparatus for distribution of data among computing resources
Publication Date: 2011.02.01 BANK OF AMERICA CORP
  • US7882200B2 patent drawing
  • US7882200B2 patent drawing
  • US7882200B2 patent drawing

AI summary

A method and apparatus for distribution of data among computing resources are disclosed. A data provider system architecture for “grid” computing resources can include a plurality of data providers. At least some of the data providers each have an associated cache for storing data to provide the data to the engines. The grid computing network interconnects the plurality of data providers and the engines. A data provider director (DPD) determines the selected data provider with respect to providing the data from the cache to a particular engine. The DPD may also provide an alternate data provider. In some embodiments, a data provider may have both an internal and external cache. Such an external cache may serve more than one data provider, for example, multiple data providers in a locale interconnected by a local area network (LAN).