TCAM Traffic Interception via Group Identifier Compression

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional ternary content-addressable memory (TCAM) systems for network security have limited capacity and are expensive, making it difficult to intercept and manage traffic from a large number of host machines effectively.

Innovation Solution

The approach involves tagging packets with group identifiers during the forwarding process, allowing a single TCAM rule to manage traffic from multiple host machines by matching on these group identifiers instead of individual hosts, thereby compressing multiple hosts into a single representation and reducing the burden on TCAM resources.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Adaptability or versatility

If TCAM is used to intercept traffic from individual host machines, then traffic interception capability is provided, but TCAM capacity is limited and memory resources are consumed quickly

Engineering Contradiction:
Improvetraffic interception capabilityVSAvoidTCAM capacity
Core Design Contradiction:
Adaptability or versatilityVSQuantity of substance

Solution Approach 1:

The patent merges multiple host machine entries into a single group entry in TCAM. Instead of storing separate rules for each host machine, the system creates a group identifier that represents multiple hosts, allowing one TCAM rule to match traffic from all hosts in the group. This combining approach directly resolves the contradiction by maintaining comprehensive traffic interception capability while dramatically reducing the number of TCAM entries needed.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The group identifier serves multiple functions: it represents multiple host machines simultaneously, enables single rule application to multiple hosts, and reduces memory consumption. This multi-functionality allows the TCAM to maintain its traffic interception capability across many host machines without requiring proportional increases in memory capacity.

Inventive Principle:
Principle #6Universality (Multi-functionality)

2Adaptability or versatility

If TCAM size is increased to handle more host machines, then more traffic can be intercepted, but cost and manufacturing complexity increase

Engineering Contradiction:
Improvenumber of host machinesVSAvoidTCAM implementation cost
Core Design Contradiction:
Adaptability or versatilityVSEase of manufacture

Solution Approach 1:

By merging multiple host entries into a single group entry, the patent reduces the total number of TCAM entries required from N (for N host machines) to 1 (for the group). This drastic reduction in entry count directly lowers the required TCAM size, making the implementation more cost-effective and easier to manufacture while maintaining the ability to handle many host machines.

Inventive Principle:
Principle #5Merging (Combining)

3Measurement precision

If individual host rules are stored in TCAM, then precise traffic targeting is achieved, but memory utilization efficiency decreases

Engineering Contradiction:
Improvetraffic targeting accuracyVSAvoidmemory utilization
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent combines multiple individual host rules into a single group rule that targets traffic from all hosts in the group simultaneously. This merging maintains precise traffic targeting capability (the system can still identify and filter traffic from specific host groups) while dramatically improving memory utilization efficiency by reducing the number of entries from N to 1.

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentUS11463479B2Intercepting network traffic
Publication Date: 2022.10.04 ARISTA NETWORKS INC
  • US11463479B2 patent drawing
  • US11463479B2 patent drawing
  • US11463479B2 patent drawing

AI summary

Incoming packets in a switch are associated with one or more group identifiers based on content contained in the incoming packets. Rules for processing the corresponding outgoing packets are identified based at least on the group identifiers associated with the incoming packets. Actions associated with matched rules are applied to the outgoing packets.