Interconnection Network Port Segmentation via Formal Language Requests

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for monitoring large interconnection networks, such as InfiniBand or OmniPath systems, are inefficient in rapidly segmenting the network into smaller subsets for distributed monitoring, requiring extensive routing tables and resource-intensive processes, which hinders quick anomaly detection and hardware failure monitoring.

Innovation Solution

A method and device that generate requests to segment the interconnection network into smaller sets of ports using formal language instructions, eliminating the need for heavy routing tables by utilizing direct route paths, allowing for targeted monitoring by dedicated software entities and reducing resource consumption.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If routing tables are used to segment the network, then network subdivision is achieved, but device complexity and resource consumption increase

Engineering Contradiction:
Improvenetwork segmentation capabilityVSAvoidrouting table complexity
Core Design Contradiction:
Adaptability or versatilityVSDevice complexity

Solution Approach 1:

The patent extracts only the essential routing information needed for segmentation (direct route paths) while eliminating the complex full routing tables. This allows network subdivision functionality to be maintained with significantly reduced data structures and computational overhead.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

Instead of using complete routing tables, the system creates simplified copies containing only the necessary path information for monitoring purposes. These condensed route representations maintain segmentation capability while reducing complexity.

Inventive Principle:
Principle #26Copying

2Reliability

If routing tables are recalculated for topology changes, then optimal performance is maintained, but processing time increases

Engineering Contradiction:
Improvemonitoring accuracyVSAvoidrouting table recalculation time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system pre-calculates and stores direct route paths during network initialization, so that when topology changes occur, only minimal updates are needed rather than complete recalculation. This preliminary preparation maintains monitoring accuracy while reducing response time to changes.

Inventive Principle:
Principle #10Preliminary action

3Reliability

If a single entity monitors the entire network, then comprehensive monitoring is achieved, but resource consumption increases

Engineering Contradiction:
Improvemonitoring coverageVSAvoidmonitoring resource consumption
Core Design Contradiction:
ReliabilityVSUse of energy by moving object

Solution Approach 1:

The patent divides the network into multiple monitorable segments using direct route paths, allowing the monitoring system to be split into multiple independent entities. Each entity monitors a specific portion of the network, achieving comprehensive coverage while distributing and reducing the resource burden on each individual monitoring unit.

Inventive Principle:
Principle #1Segmentation

4Measurement precision

If direct route commands are used for each port, then targeted monitoring is achieved, but the number of commands becomes excessively large

Engineering Contradiction:
Improvemonitoring target precisionVSAvoidcommand generation efficiency
Core Design Contradiction:
Measurement precisionVSProductivity

Solution Approach 1:

The patent merges multiple individual direct route commands into a single aggregated request that targets multiple ports simultaneously. This consolidation maintains the precision of targeted monitoring while dramatically improving command generation efficiency and reducing the overhead of managing thousands of separate commands.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP3471346B1Method for generation of requests for the segmentation of monitoring of an interconnection network and associated equipment
Publication Date: 2022.11.23 BULL SA
  • EP3471346B1 patent drawingFigure 1
  • EP3471346B1 patent drawingFigure 2
  • EP3471346B1 patent drawingFigure 3A

AI summary

The invention relates to a method for generating a request, from a formal language instruction defining a set of ports of an interconnection network, said request comprising an addressing command for each of the ports defined in the instruction, said method comprising the following steps: - Reception, by a communication module, of a formal language instruction defining a set of ports, - Processing, by a processing module, of the formal language instruction so as to generate a set of numbers coded on at least one byte, each number comprising position bits, each of the position bits allowing to identify a port and at least one authorization bit, the at least one authorization bit allowing to define access rights on the ports, and - Encoding, by an encoding module, of the set of numbers so as to generate the request comprising the addressing command.