Instantaneous Packet Failover via Primary-Secondary Flow Assignment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional failover processes in packet processing systems are time-consuming, leading to packet loss or drop during the failover response period, as they rely on external layer 2 switches and gratuitous ARP messages to redirect traffic, resulting in delays of several seconds.

Innovation Solution

A system and method for instantaneous failover, where a packet distribution module maintains information on the operating status of packet processing elements and assigns primary and secondary elements for each packet flow, allowing for immediate routing based on operational status, eliminating the need for gratuitous ARP messages and reducing failover response time.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If conventional failover process using external layer 2 switch and gratuitous ARP messages is used, then traffic can be redirected from primary to secondary chassis, but failover response time increases to multiple seconds causing packet loss

Engineering Contradiction:
Improvefailover capabilityVSAvoidfailover response period
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent extracts the load balancing function from the external layer 2 switch and relocates it directly into the packet processing elements themselves. Each packet processing element maintains local load balancing information and can independently make routing decisions, eliminating the need for external switch intervention and ARP message exchanges during failover.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent implements preliminary action by pre-configuring secondary packet processing elements as backups before any failure occurs. The load balancing module maintains up-to-date information about which elements are primary and which are secondary for each packet flow, so that when a failure occurs, the switch can immediately redirect traffic without waiting for failure detection or ARP messages.

Inventive Principle:
Principle #10Preliminary action

2Reliability

If external layer 2 switch is used to effect failover by updating routing tables, then traffic redirection is achieved, but the process takes multiple seconds during which packets are lost

Engineering Contradiction:
Improvetraffic redirectionVSAvoidpacket processing continuity
Core Design Contradiction:
ReliabilityVSProductivity

Solution Approach 1:

The packet processing elements perform self-service by maintaining their own load balancing information and failure state knowledge locally. When a failure occurs, the affected packet processing element can immediately begin redirecting its own traffic without requiring external switch intervention, enabling instantaneous failover and maintaining packet processing continuity.

Inventive Principle:
Principle #25Self-service

3Reliability

If gratuitous ARP messages are used to notify layer 2 switch of primary failure, then the switch can update its routing table, but this notification process introduces multiple seconds of delay

Engineering Contradiction:
Improvefailure notificationVSAvoidnotification delay
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent removes the ARP notification mechanism entirely from the failover process. Instead of using ARP messages to notify the switch of failures, the system uses direct internal communication between packet processing elements and the load balancing module, which maintains real-time knowledge of element status and can immediately redirect traffic without any ARP message delays.

Inventive Principle:
Principle #2Taking out (Extraction)

Data Source

PatentUS8472311B2Systems, methods, and computer readable media for providing instantaneous failover of packet processing elements in a network
Publication Date: 2013.06.25 RIBBON COMMUNICATIONS OPERATING CO INC
  • US8472311B2 patent drawing
  • US8472311B2 patent drawing
  • US8472311B2 patent drawing

AI summary

Systems, methods, and computer program products for providing instantaneous failover of packet processing elements in a network are disclosed. According to one aspect, the subject matter described herein includes a system for providing instantaneous failover of packet processing elements in a network. The system includes a plurality of packet processing elements for processing packets in a network and a packet distribution module for maintaining information about the operating status of each element of the plurality of packet processing elements, for maintaining information about packet flows being processed by the packet processing elements, and, for each packet flow, assigning one of the plurality of packet processing elements as the primary element for the packet flow and assigning another of the plurality of packet processing elements as the secondary element for the packet flow. The packet distribution module routing packets to the packet processing elements according to the operating status of the packet processing elements that have been assigned to the packet flow with which the packets are associated.