DDS QoS Manager for HPC Partitioning

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing distributed memory HPC and grid computing systems lack proper quality of service (QoS) control, leading to inefficiencies in data processing and increased time consumption due to communication latency and bandwidth unpredictability, resulting in wasted processing time and resource duplication when handling large datasets for hydrocarbon exploration and production.

Innovation Solution

A data processing system with master nodes acting as publishers and processor nodes as subscribers, utilizing a quality of service standard profile to manage data partitions and ensure compliance with established QoS settings, implementing the Data Distribution Service (DDS) standard for scalable and reliable data distribution, enabling fault-tolerant processing and efficient resource allocation.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If standard communication libraries (MPI, PVM) are used for data distribution, then device complexity is reduced, but quality of service control capability is lost

Engineering Contradiction:

Improvecommunication library complexityVSAvoidquality of service control

Core Design Contradiction:

Device complexityVSReliability

Solution Approach 1:

The patent introduces a Quality of Service Manager as an intermediary component that mediates between the data distribution system and the communication infrastructure. This manager enables QoS control by monitoring communication parameters and adjusting data distribution accordingly, resolving the contradiction between simple communication libraries and reliable QoS control.

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces the mechanical communication library interface with a software-based QoS management mechanism. Instead of relying on fixed MPI/PVM calls, the system uses dynamic QoS parameters and adaptive routing to achieve reliable service quality control while maintaining programming simplicity.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

2Speed

If high-speed interconnects (Infiniband, Myrinet, Quadrics, Gigabit Ethernet) are used for data communication, then processing speed is improved, but communication latency and bandwidth predictability deteriorate

Engineering Contradiction:

Improvedata communication speedVSAvoidcommunication latency and bandwidth predictability

Core Design Contradiction:

SpeedVSReliability

Solution Approach 1:

The QoS Manager continuously monitors communication metrics including latency and bandwidth utilization from high-speed interconnects. Based on this feedback, the system dynamically adjusts data distribution decisions to compensate for unpredictable performance variations, ensuring consistent QoS despite the inherent variability of high-speed networks.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent implements dynamic adaptation to network conditions by adjusting communication parameters in real-time. The system can modify data distribution strategies based on current network state, transforming the static communication patterns into dynamic responses that maintain predictability despite high-speed network variability.

Inventive Principle:

Principle #15Dynamics

3Device complexity

If data processing is performed without QoS control, then device complexity is reduced, but processing time increases due to inefficiencies and resource wastage

Engineering Contradiction:

Improvesystem control complexityVSAvoidprocessing time

Core Design Contradiction:

Device complexityVSLoss of time

Solution Approach 1:

The QoS Manager performs preliminary assessment of data priorities and communication requirements before data distribution begins. By pre-establishing QoS parameters and prioritization rules, the system avoids time-consuming adjustments during processing, reducing overall processing time without significantly increasing operational complexity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system changes key parameters such as data priority levels, transmission timing, and resource allocation based on QoS requirements. These parameter adjustments enable efficient processing by directing computational resources to critical operations first, reducing total processing time while maintaining manageable system complexity through automated parameter management.

Inventive Principle:

Principle #35Parameter changes

Data Source

PatentUS9482769B2High performance and grid computing with partitioning quality of service control

Publication Date: 2016.11.01 SAUDI ARABIAN OIL CO

AI summary

High performance computing (HPC) and grid computing processing for seismic and reservoir simulation are performed without impacting or losing processing time in case of failures. A Data Distribution Service (DDS) standard is implemented in High Performance Computing (HPC) and grid computing platforms, to avoid the shortcomings of current Message Passing Interface (MPI) communication between computing modules, and provide quality of service (QoS) for such applications. QoS properties of the processing can be controlled. A “partitioning” quality of service is provided and the computer can be logically segregated into several “logical partitions” so that a computer can have several publisher nodes, serving several groups of compute nodes, and running different applications independently.