Network Flow Deduplication via Hash-Based Cluster Coordination

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

In distributed network monitoring environments, multiple network monitoring devices often receive duplicate network traffic, making it difficult to determine correlations and extract meaningful information from duplicative data, leading to challenges in troubleshooting and analysis.

Innovation Solution

Implementing a system where network monitoring computers (NMCs) in a cluster register and manage network flows using hashing and policy-based methods to identify responsible NMCs for processing specific flows, thereby eliminating duplicates and optimizing data processing.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If multiple network monitoring devices are deployed to monitor network traffic, then monitoring coverage and reliability are improved, but data duplication increases and processing efficiency deteriorates

Engineering Contradiction:
Improvemonitoring reliabilityVSAvoiddata processing efficiency
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The patent divides the network monitoring task into segments by assigning different network flows to different NMCs based on hashing of flow identifiers. Each NMC monitors specific flows rather than all traffic, eliminating duplicates while maintaining comprehensive coverage through coordinated segmentation of the monitoring workload

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces a flow manager as an intermediary component that coordinates between multiple NMCs. The flow manager receives flow identification information, determines responsible NMCs using hashing, and directs traffic accordingly, enabling efficient deduplication and load distribution across the monitoring cluster

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If multiple network monitoring devices process all network traffic, then monitoring accuracy is improved, but resource consumption and system complexity increase

Engineering Contradiction:
Improvemonitoring accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent implements local quality by having each NMC specialize in monitoring specific network flows determined by hashing. Each device optimizes its resources for its assigned flows rather than processing all traffic, reducing individual device complexity while maintaining overall monitoring accuracy through the collective capability of the cluster

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent changes the parameter of flow assignment from static to dynamic based on hashing calculations. The system uses configurable hash functions and policies to determine flow distribution, allowing flexible adjustment of monitoring parameters while maintaining consistent deduplication behavior across the cluster

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentUS9660879B1Flow deduplication across a cluster of network monitoring devices
Publication Date: 2017.05.23 EXTRAHOP NETWORKS INC
  • US9660879B1 patent drawing
  • US9660879B1 patent drawing
  • US9660879B1 patent drawing

AI summary

Embodiments are directed to monitoring flows of packets over a network. If a network monitoring computer (NMC) in a cluster of NMCs observes a new network flow, the NMC may perform a variety of actions to determine the NMC that is responsible for monitoring the new network flow. Network traffic associated with the new network flow may be buffered in a non-transitory processor readable media. The new network flow may be registered with the plurality of NMCs, providing an identifier that corresponds to one NMC. Registering may include, assigning the NMC a responsibility to monitor the new network flow. If the identifier corresponds to the NMC that observed the new network flow, the network traffic associated with the new network flow is processed using that NMC. If the identifier corresponds to another NMC, the buffered network traffic is forwarded to the other NMC.