Container Traffic Identification via Network Layer Extraction

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current technologies lack the ability to uniquely identify and isolate container-based workloads on the network layer, preventing effective policy enforcement and management of container traffic, as IP addresses and ports are insufficient for distinguishing between containers offering different services on the same compute node or across nodes.

Innovation Solution

A system and method that extracts container identification data from packet data fields, such as IPv6 extension headers or Network Service Headers, to apply policies on a per-container or per-container-flow basis, enabling QoS, prioritization, accounting, and cloud-based classification.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Device complexity

If IP addresses and TCP/UDP ports are used for identification, then network communication is simplified, but the ability to uniquely identify and isolate container traffic is lost

Engineering Contradiction:
Improvenetwork identification complexityVSAvoidcontainer traffic identification accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the network identification space by introducing a unique container identifier (such as a modified TCP port field or custom header) that divides traffic identification into two levels: traditional IP/port for host-level routing and container-specific identifiers for container-level isolation. This segmentation allows each container to be uniquely identified even when sharing the same compute node and application port.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent adds another dimension to network identification by embedding container-specific identifiers within existing packet structures (such as using the TCP port field or adding custom headers). This dimensional addition enables differentiation of container traffic without fundamentally changing the underlying IP addressing scheme, thus maintaining compatibility while improving identification precision.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Productivity

If containers share the same compute node and application port, then resource utilization is improved, but the ability to apply per-container policies is lost

Engineering Contradiction:
Improveresource utilization efficiencyVSAvoidper-container policy enforcement capability
Core Design Contradiction:
ProductivityVSAdaptability or versatility

Solution Approach 1:

The patent applies local quality by associating specific identification markers with individual container traffic flows. Each container's traffic is tagged with its unique identifier at the network layer, allowing network components to apply differentiated policies (such as QoS, routing, or security rules) to specific containers while they share the same physical infrastructure and application ports.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent introduces an intermediary identification mechanism (container-specific identifier embedded in packet headers) that mediates between the shared resource environment and the need for individualized policy control. This intermediary allows the network to distinguish and control individual container traffic without requiring separate physical or application-level resources.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Reliability

If container identification is implemented at the network layer, then policy enforcement capability is improved, but packet processing complexity increases

Engineering Contradiction:
Improvepolicy enforcement effectivenessVSAvoidpacket processing complexity
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent applies preliminary action by embedding container identification information in packet headers at the source (network layer) before packets traverse the network. This pre-tagging approach allows intermediate network devices to quickly identify and apply appropriate policies without performing complex analysis or lookups during packet processing, thus reducing processing complexity while maintaining enforcement effectiveness.

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP3488583B1System and method for transport-layer level identification and isolation of container traffic
Publication Date: 2021.09.01 CISCO TECHNOLOGY INC
  • EP3488583B1 patent drawingFigure 1
  • EP3488583B1 patent drawingFigure 2
  • EP3488583B1 patent drawingFigure 3

AI summary

Disclosed is a system and method of providing transport-level identification and isolation of container traffic. The method includes receiving, at a component in a network, a packet having a data field, extracting, at a network layer, container identification data from the data field and applying a policy to the packet at the component based on the container identification data. The data field can include one of a header, an IPv6 extension header, a service function chaining container identification, a network service header, and an optional field of an IPv4 packet.