Layer 3 VM Mobility Protocol for Datacenter Scalability
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Datacenter networks, particularly those implemented in layer 2, face limitations in scaling for address resolution protocols and lack a control plane protocol for OSI layer 3-based virtual machine (VM) mobility, hindering efficient VM movement and load balancing.
Innovation Solution
A layer 3 VM mobility protocol and architecture are introduced, where a hypervisor registers VMs with a management node, updating binding caches and using Proxy-COA to manage VM mobility across hypervisors, supporting both IPv4 and IPv6, and optimizing routing through interior and inter-domain protocols.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If datacenter networks are implemented in layer 2, then VM mobility can be supported, but scalability is limited due to address resolution protocol constraints
Solution Approach 1:
The patent introduces a Proxy Mobile IPv6 gateway as an intermediary between layer 2 VM mobility mechanisms and layer 3 network protocols. This gateway translates VM mobility requests into Proxy-CoA updates, enabling VM migration across layer 3 boundaries without direct layer 2 connectivity, thus resolving the scalability limitation while maintaining VM mobility support
Solution Approach 2:
The patent transitions VM mobility from a layer 2 constraint to a layer 3 solution by introducing IP-based mobility management. By using Proxy Mobile IPv6 and binding caches at the network layer, the system enables VM migration across different network segments and data centers, adding a new dimensional approach to VM mobility that overcomes layer 2 scalability limits
2Ease of manufacture
If VM mobility is implemented without a control plane protocol, then implementation is simpler, but VM movement and load balancing efficiency are hindered
Solution Approach 1:
The patent segments the VM mobility control plane into distinct components: management nodes for binding cache management, proxy gateways for mobility signaling, and hypervisors for VM migration execution. This segmented architecture provides structured control for efficient VM movement while maintaining modular implementation that balances complexity and efficiency
3Loss of energy
If VM migration is performed without binding cache updates, then signaling overhead is reduced, but communication continuity cannot be maintained
Solution Approach 1:
The patent implements preliminary binding cache updates at the proxy gateway before VM migration completes. By pre-establishing the Proxy-CoA binding in the binding cache, the system ensures that network traffic can be immediately redirected to the new VM location, maintaining communication continuity while minimizing signaling overhead through efficient cache utilization
Data Source
AI summary
A network element (NE) comprising a hypervisor configured to receive a first virtual machine (VM) from a source hypervisor, and send a first VM mobility registration request to a management node, wherein the first VM mobility registration request comprises a VM identifier associated with the first VM and an internet protocol (IP) address associated with the hypervisor. Also disclosed is a management node comprising a binding cache, and a processor, wherein the processor is configured to receive a first VM mobility registration request from a first hypervisor, wherein the first VM mobility registration is associated with a VM, and record VM information from the first VM mobility registration request in an entry of the binding cache.