Cloud Failover Routing Using Virtual IP and Packet Metadata
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing cloud computing environments face challenges in handling device switchovers due to distributed architecture, resource elasticity, diverse workloads, and latency, particularly when using private IP addresses and devices that lack support for protocols like HSRP and VRRP.
Innovation Solution
Implementing a method that generates a virtual routing IP address common to multiple network nodes, sending the physical IP address as metadata in packets to upstream devices, and using AI and machine learning to predict failures and update key-value mappings for seamless failovers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional hardware redundancy and manual intervention methods are used to handle device failures, then device switchover can be achieved, but the system is not scalable and inefficient in cloud computing environments
Solution Approach 1:
The patent replaces manual hardware intervention with automated software-based failover mechanisms. The system automatically detects device failures and transfers operations to backup systems without human intervention, transforming the mechanical/manual switchover process into an automated computational system that is scalable to cloud environments
Solution Approach 2:
The patent creates a universal failover system that can handle multiple types of devices and failure scenarios through a single automated platform. The key-value store architecture allows the system to manage failover for various network devices (layer 3 devices, virtual devices) using the same mechanism, enhancing both reliability and scalability
2Reliability
If automated failover mechanisms are implemented to maintain high availability, then service continuity is improved, but network elements may not properly react to switchovers and route data packets to correct devices
Solution Approach 1:
The patent implements a feedback mechanism where the system continuously monitors device status and automatically updates network routing information. When a failover occurs, the system detects the change and notifies upstream network elements, ensuring they have current routing information to direct data packets to the correct active device
Solution Approach 2:
The patent introduces an intermediary key-value store that acts as a mediator between the failover system and upstream network elements. This intermediary stores and manages routing information, allowing network elements to query the current active device and receive accurate routing updates without direct complex communication protocols
3Reliability
If private IP addresses are used in distributed cloud architecture, then network security and isolation are improved, but protocols like HSRP and VRRP cannot be supported and failover becomes more complex
Solution Approach 1:
The patent uses a key-value store to copy and manage routing information across the distributed system. Instead of relying on complex protocol mechanisms, the system maintains simplified routing data in the key-value store that can be queried by network elements, reducing failover complexity while maintaining security through private IP addressing
Solution Approach 2:
The patent changes the approach from protocol-based failover (HSRP/VRRP) to a parameter-based system using key-value stores. By storing routing information as simple key-value pairs rather than relying on complex protocol parameters, the system achieves failover capability with private IPs without the complexity of traditional protocols
Data Source
AI summary
The present application discloses a method, system, and computer system for performing failovers of traffic carrying devices. The method includes (i) generating, by one or more processors a virtual routing IP address that is common to a plurality of network nodes, and (ii) sending to an upstream device a physical IP address for a particular network node of the plurality of network nodes, wherein the physical IP address is sent as metadata in a packet to the upstream device.