Job Scheduling Manager for HPC Load Balancing

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In high-performance computing (HPC) environments, users face challenges in controlling resource distribution and job scheduling without manual text-based commands, limiting their ability to efficiently direct processor-heavy and memory-heavy jobs to appropriate nodes.

Innovation Solution

A system and method that includes a processor and memory to receive user jobs and preferences, automatically generating compatible commands for the load balancer, allowing users to transmit jobs and preferences seamlessly, and notifying users of successful transmission, thereby enabling efficient workload distribution and job scheduling through a graphical user interface.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If manual text-based commands are used to control load balancer, then users can direct jobs to appropriate nodes, but the operation complexity and time consumption increase

Engineering Contradiction:
Improveease of job submissionVSAvoidtime for command input
Core Design Contradiction:
Ease of operationVSLoss of time

Solution Approach 1:

The patent introduces an intermediary system (job scheduling interface with automated command generation) that mediates between the user and the load balancer. Users submit jobs through a simplified interface, and the system automatically generates and transmits the appropriate text-based commands to the load balancer, eliminating the need for users to manually input complex commands while maintaining full control over job distribution

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements self-service by automatically generating the necessary load balancer commands based on user-submitted job parameters. The automated command generation system analyzes job requirements and constructs appropriate text-based commands without human intervention, allowing the system to serve itself in translating high-level user intentions into low-level load balancer instructions

Inventive Principle:
Principle #25Self-service

2Ease of operation

If automated command generation is implemented, then user interaction is simplified, but the system complexity increases

Engineering Contradiction:
Improveease of job submissionVSAvoidsystem complexity
Core Design Contradiction:
Ease of operationVSDevice complexity

Solution Approach 1:

The patent introduces an intermediary system (job scheduling interface with automated command generation) that mediates between the user and the load balancer. Users submit jobs through a simplified interface, and the system automatically generates and transmits the appropriate text-based commands to the load balancer, eliminating the need for users to manually input complex commands while maintaining full control over job distribution

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The system implements self-service by automatically generating the necessary load balancer commands based on user-submitted job parameters. The automated command generation system analyzes job requirements and constructs appropriate text-based commands without human intervention, allowing the system to serve itself in translating high-level user intentions into low-level load balancer instructions

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS10671433B2Load balancing and job scheduling manager within a high-performance computing environment
Publication Date: 2020.06.02 TATA CONSULTANCY SERVICES LTD
  • US10671433B2 patent drawing
  • US10671433B2 patent drawing
  • US10671433B2 patent drawing

AI summary

This disclosure relates generally to systems and methods for load balancing, resource allocation, and job scheduling within a high-performance computing (HPC) environment. In one implementation, the system may include a processor configured to execute instructions and a memory storing the instructions. Furthermore, the instructions may include commands to receive at least one job from a user; receive at least one preference from a user; automatically generate, in a language compatible with the load balancer, at least one command based on the at least one job and the at least one preference; transmit the at least one job and the at least one command to the load balancer; and notify the user that the received at least one job and the received at least one preference have been transmitted to the load balancer.