Independent VNF Component Clustering for NFV Scaling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Network Function Virtualization (NFV) faces challenges in providing high availability and elasticity due to the failure of virtual machines and physical machines that support them, as existing redundancy mechanisms and scaling strategies are not optimized for different VNF components with distinct functions.
Innovation Solution
The solution involves logically grouping VNF components into clusters (service, database, and load balancing) with independent scaling rules and redundancy models, allowing for dynamic provisioning and decommissioning of virtual machines based on demand and resource utilization, and using active/active or active/standby redundancy models to ensure high availability.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If VNF components are managed as a single homogeneous group, then management complexity is reduced, but scaling efficiency and resource optimization deteriorate
Solution Approach 1:
The patent divides VNF components into distinct clusters (service cluster, database cluster, load balancing cluster) based on their functional characteristics. Each cluster is managed independently with its own scaling policies and redundancy models, allowing optimized resource allocation for each component type while maintaining overall system manageability through automated cluster-level control.
2Device complexity
If uniform redundancy models are applied to all VNF components, then system configuration simplicity is maintained, but availability and fault tolerance deteriorate
Solution Approach 1:
The patent applies different redundancy models to different VNF clusters based on their specific availability requirements. Service clusters use active/active redundancy for high availability, database clusters use active/standby for data persistence, and load balancing clusters use active/active for continuous service. This localized approach optimizes reliability for each component type while the NFV manager provides centralized coordination.
3Productivity
If VNF components scale independently, then resource optimization and elasticity improve, but coordination complexity and system stability deteriorate
Solution Approach 1:
The patent implements automated scaling mechanisms where each cluster monitors its own resource utilization and demand metrics. The NFV manager receives feedback from all clusters and coordinates scaling actions across clusters to maintain system stability. This feedback loop enables independent resource optimization while the centralized manager ensures coherent system-wide scaling decisions.
4Adaptability or versatility
If dynamic scaling is implemented for VNF clusters, then elasticity and demand responsiveness improve, but system stability and coordination reliability deteriorate
Solution Approach 1:
The patent enables dynamic scaling of VNF clusters based on real-time demand metrics and resource utilization patterns. The NFV manager continuously monitors cluster performance and automatically adjusts cluster sizes and configurations to match actual service demands. This dynamic approach provides elasticity while the manager's coordinated control maintains system stability during scaling transitions.
Data Source
AI summary
A method includes executing a Virtual Network Function (VNF) that includes a plurality of VNF components supported by a plurality of virtual machines, the virtual machines supported by a set of physical machines, the plurality of VNF components comprising a first group of VNF components and a second group of VNF components that is different than the first group, both the first group and the second group being independently scalable. The method further includes scaling the first group of VNF components in response to a change in demand for services associated with the first group.