Self-Managed Virtual Network Architecture for High Availability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
High availability designs for virtualized systems and services are complex due to multiple independent layers with their own failure recovery mechanisms, making it challenging to achieve near 100% operation time, as existing solutions often focus on single or dual layers and lack end-to-end service level coordination.
Innovation Solution
Implementing a self-managed network architecture with a centralized or distributed management system using artificial intelligence and analytics, where virtual network elements and virtual network functions can self-configure, diagnose, and repair issues, with standardized message formats for failure reporting and coordinated failure recovery timers to prevent race conditions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple independent layers with their own failure recovery mechanisms are implemented, then each layer can be managed independently, but the overall system complexity increases significantly
Solution Approach 1:
An orchestrator is introduced as an intermediary component that coordinates failure recovery across multiple layers. The orchestrator receives failure notifications from any layer, determines the appropriate recovery action, and manages the recovery process centrally, thereby reducing the complexity of inter-layer coordination while preserving independent layer management capabilities
2Reliability
If each layer has its own failure recovery mechanism, then layer-specific issues can be addressed, but coordination among layers becomes difficult
Solution Approach 1:
A feedback mechanism is implemented where layers send failure notifications to the orchestrator, which then coordinates the recovery process and provides status updates back to the affected layers. This feedback loop enables automated coordination among layers while maintaining layer-specific recovery capabilities, as each layer can independently detect failures and report them to the central orchestrator for coordinated response
3Productivity
If standardized message formats are implemented for failure reporting, then communication efficiency improves, but implementation complexity increases
Solution Approach 1:
A universal standardized message format is designed that can be used across all layers and by the orchestrator for failure reporting and coordination. This single standardized format serves multiple functions: it enables consistent communication between any layer and the orchestrator, simplifies the orchestrator's processing logic, and reduces implementation complexity by providing a common interface that all components must follow, thereby improving overall communication efficiency
Data Source
AI summary
Systems and methods described herein include high-availability layers for virtualized systems and services. Virtual network elements include an infrastructure layer that automatically detects and repairs failures in hardware and an operating system for one of the virtual network elements; a virtualization layer that automatically detects and repairs failures in a hypervisor executing on the hardware and the operating system; a virtual machine layer that automatically detects and repairs failures in virtual machines managed by the hypervisor; a virtual network function layer that automatically detects and repairs failures in virtual network functions executed on the virtual machines; and a connection layer that automatically detects and repairs failures in connections between the virtual network elements. Each layer includes a fix agent, which uses artificial intelligence rules to automatically provide an estimated fix time interval for a local failure and to repair the local failure, and a fault management agent, which diagnoses failures within the corresponding layer.