Transparent Network Services Elastic Scale-Out via ECMP
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing network service capacity scaling methods often disrupt existing traffic flows and lead to inefficient distribution of traffic, as individual boxes handle traffic independently, making it difficult to maintain connections and scale capacity transparently to client devices.
Innovation Solution
A distributed network service platform that configures additional devices to provide network services during a scale-out process, utilizing Equal-Cost Multi-Path (ECMP) routing or alternative load balancing techniques to maintain transparent scaling and efficient traffic distribution, ensuring existing connections are maintained and new flows are balanced across devices.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If individual boxes handle traffic independently for network services, then device complexity is reduced and ease of operation is improved, but capacity scaling interrupts existing traffic flows and leads to inefficient distribution of traffic flows
Solution Approach 1:
The patent merges multiple independent network service boxes into a unified distributed platform where devices cooperate through ECMP routing. Instead of isolated boxes handling traffic independently, the system combines their capabilities while maintaining coordinated traffic distribution, allowing scale-out without interrupting existing flows.
Solution Approach 2:
The system implements dynamic traffic flow management where the routing behavior adapts based on flow state. New flows are distributed across multiple devices using ECMP, while existing flows are pinned to specific devices to prevent interruption during scaling operations.
2Productivity
If capacity is scaled out by adding more devices, then network service capacity is improved, but traffic flow distribution becomes inefficient and connections may be interrupted
Solution Approach 1:
The patent segments traffic flow handling into two distinct mechanisms: flow pinning for existing connections to maintain continuity, and ECMP routing for new flows to enable efficient distribution across scaled-out devices. This segmentation allows capacity expansion without compromising connection stability.
Solution Approach 2:
The system introduces an intermediary routing mechanism that sits between the scaled-out devices and the traffic flows. This intermediary layer (implemented through ECMP routing tables and flow pinning logic) coordinates traffic distribution, ensuring that adding capacity doesn't disrupt existing connections while efficiently balancing new flows.
3Productivity
If ECMP routing is used for load balancing across multiple devices, then traffic flow distribution efficiency is improved, but device complexity increases
Solution Approach 1:
The patent makes ECMP routing a universal solution that handles both load balancing for new flows and coordinated scaling for multiple devices. Rather than implementing separate mechanisms for different scenarios, the system uses ECMP routing as a multi-functional foundation that simplifies overall system complexity despite the advanced routing capabilities.
Data Source
AI summary
In a network with at least a first device already configured to provide a network service to a network application, scaling service capacity includes: configuring one or more second devices to provide the network service to the network application. In embodiments where an upstream network device supports Equal-Cost Multi-Path (ECMP) routing, the upstream network device is configured, including storing a plurality of paths to reach an address associated with a network application, wherein the plurality of paths are equal in cost. In embodiments where the upstream network device does not support ECMP routing, the second device is configured not to respond to an Address Resolution Protocol (ARP) request associated with an Internet Protocol (IP) address of the network application, and the first device is instructed to perform load balancing on network traffic destined for the network application among the first device and the one or more second devices.