ARP Cache Refresh for Data Center Host Mobility

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing solutions for maintaining communication between data centers during host mobility, such as ARP cache refresh, face challenges like traffic tromboning, increased latency, and the need for explicit L2 extensions or virtual switches, which are resource-intensive and not always feasible.

Innovation Solution

Implementing an ARP/NDP refresh logic within network elements or distributed across them to identify host movements and update ARP or NDP caches by snooping packets and sending gratuitous/unsolicited messages to ensure accurate data link layer addresses for communication continuity across data centers.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If ARP cache refresh is implemented during host mobility, then communication continuity is improved, but traffic latency increases due to tromboning

Engineering Contradiction:
Improvecommunication continuityVSAvoidtraffic latency
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent introduces a virtual switch as an intermediary component that maintains ARP cache bindings and intercepts traffic during host mobility transitions. This virtual switch acts as a mediator between the moving host and the network, allowing ARP cache updates without forcing traffic through distant paths, thereby reducing latency while maintaining communication continuity.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent implements preliminary ARP cache refresh actions before the host actually moves between data centers. By proactively updating ARP bindings in advance and maintaining them through the transition period, the system prevents traffic black-holes and reduces the need for reactive traffic rerouting, thereby minimizing latency while ensuring communication continuity.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If L2 extensions or virtual switches are deployed for ARP cache refresh, then communication during mobility is improved, but device complexity and resource consumption increase

Engineering Contradiction:
Improvecommunication during mobilityVSAvoidL2 extensions and virtual switches
Core Design Contradiction:
ReliabilityVSDevice complexity

Solution Approach 1:

The patent makes existing network elements (routers, switches, firewalls) multi-functional by enabling them to perform ARP cache refresh operations natively. Instead of adding dedicated L2 extension infrastructure or separate virtual switch components, the patent allows these universal network elements to handle both their traditional functions and ARP cache management, thereby reducing device complexity while maintaining communication reliability during mobility.

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

3Device complexity

If ARP cache is not refreshed during host mobility, then device complexity is reduced, but communication failures occur due to traffic black-holes

Engineering Contradiction:
ImproveARP cache managementVSAvoidcommunication reliability
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent implements self-service ARP cache refresh where network elements automatically detect host mobility events and trigger appropriate ARP cache updates without external intervention. The system monitors traffic patterns, identifies mobility events, and autonomously refreshes ARP bindings, thereby maintaining communication reliability while keeping device complexity low through automated self-management rather than manual or externally-controlled processes.

Inventive Principle:
Principle #25Self-service

Data Source

PatentUS20240323091A1Refresh of the binding tables between data-link-layer and network-layer addresses on mobility in a data center environment
Publication Date: 2024.09.26 CISCO TECHNOLOGY INC
  • US20240323091A1 patent drawing
  • US20240323091A1 patent drawing
  • US20240323091A1 patent drawing

AI summary

A method for assisting communication of a source host upon movement from a first Data center (DC) to a second DC is disclosed. The method includes identifying that the source host has moved from the first DC to the second DC, ensuring that packets identifying a source as the source host in the second DC are copied to a control plane network element, and, for a first destination host identified in a first packet copied to the control plane network element and identified as a host that is not in the second DC, updating an Address Resolution Protocol (ARP)/Neighbor Discovery Protocol (NDP) cache of the source host by sending, to the source host, a first ARP message/unsolicited neighbor advertisement specifying a Media Access Control (MAC) address of an edge router associated with the source host in the second DC as a destination MAC address for the first destination host.