Grid Computing Management Server Load Distribution
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In grid computing systems, as the number of computers increases, the load on the job requester's server also increases, leading to concentrated network traffic and the need for costly server upgrades to efficiently process large jobs.
Innovation Solution
A grid computing system where multiple processing servers execute assigned programs with input data, with a management server generating execution directions and sending input data to the first processing server, which then transfers updated data to subsequent servers, reducing the load on the management server and distributing processing tasks across multiple computers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If the job requestor's server sends programs and data to multiple computers for parallel processing, then job execution efficiency is improved, but server load and network traffic concentration increase
Solution Approach 1:
The patent segments the job processing function across multiple processing servers instead of concentrating it on a single requestor's server. Each processing server independently receives job requests, manages local program execution, and handles data processing, thereby distributing the computational load and reducing the burden on any single server while maintaining parallel processing efficiency.
2Productivity
If more computers are used to process jobs in parallel, then processing capacity increases, but network traffic concentration on the requestor's server increases
Solution Approach 1:
The patent introduces processing servers as intermediary nodes between the job requestor and the computational resources. These intermediaries receive job requests, distribute tasks to appropriate processing servers, and coordinate data flow, thereby reducing direct network traffic between the requestor and multiple processing nodes while enabling efficient parallel processing through the intermediary layer.
Data Source
AI summary
A method and system for performing a requested job. The system includes multiple processing servers and a management server. Each processing server executes an assigned program for performing the requested job. An execution direction, which identifies each of the multiple programs and an execution order for their execution (including identification of a first program to be executed), is generated. The management server sends to a first processing server the execution direction and input data. The first processing server: executes the first program using the input data, resulting in updating the input data; sends to the management server an inquiry for identification information that identifies a second processing server to execute a second program included in the execution direction; receives the identification information from the management server; and sends to the second processing server the execution direction and the updated input data for subsequent execution of the second program.