VM Migration via Information-Centric Networking
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for virtual machine (VM) migration in datacenter networks face scalability issues and complexity when dealing with large numbers of servers and routers, particularly in maintaining unchanged IP/MAC addresses across migrations, which complicates client service continuity and network resource management.
Innovation Solution
A system and method for VM migration that uses a VM mobility manager to maintain globally unique VM identifiers, enabling seamless migration by routing data packets based on VM identifiers and updating location information, while utilizing hypervisors and virtual switches to manage shareable and non-shareable resources, and employing information-centric networking for decentralized control and storage.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional VM migration methods are used with centralized storage and dedicated networks, then VM mobility is achieved, but scalability deteriorates with increasing numbers of servers and routers
Solution Approach 1:
The patent segments the centralized storage system into distributed storage nodes across multiple datacenters. Each datacenter maintains local storage capabilities, eliminating the single point of failure and scaling limitations of centralized storage. This segmentation allows the system to handle increasing numbers of servers and routers without proportionally increasing network complexity.
Solution Approach 2:
The patent introduces an information-centric networking layer that acts as an intermediary between VMs and physical infrastructure. This layer uses content-based routing and named data structures to abstract the underlying network complexity, enabling VM migration without requiring complex network reconfiguration or dedicated migration networks.
2Reliability
If VM migration maintains unchanged IP/MAC addresses, then client service continuity is achieved, but routing complexity increases across datacenters
Solution Approach 1:
The patent adds a new dimension to VM identification by using globally unique VM names in addition to traditional IP/MAC addresses. This naming dimension enables content-based routing where packets are routed based on the VM's identity rather than its location, maintaining service continuity while simplifying routing across distributed datacenters.
Solution Approach 2:
The patent creates a universal naming scheme that works across multiple datacenters and network domains. The globally unique VM name serves multiple functions: identification, routing, and location tracking, eliminating the need for complex address translation and routing protocols required by traditional methods.
3Ease of manufacture
If centralized storage capabilities are used for VM migration, then migration is enabled, but scalability is limited with increasing server numbers
Solution Approach 1:
The patent divides centralized storage into distributed storage nodes located across multiple datacenters. Each node can independently store and retrieve VM images and data, enabling parallel operations and eliminating the bottleneck of centralized storage. This segmentation directly improves system scalability while maintaining migration capability.
Solution Approach 2:
The patent transitions from location-based storage access to content-based storage access using information-centric networking principles. Storage resources are accessed through named references rather than physical locations, enabling flexible data movement and improving system scalability without sacrificing migration capability.
Data Source
AI summary
An apparatus comprising a memory and a processor coupled to the memory, wherein the memory contains instructions that when executed by the processor cause the apparatus to receive a migration request from a virtual machine (VM) proxy running on a first server hosting a VM, wherein the migration request comprises a VM identifier associated with the VM comprising a globally unique VM name and a current VM address associated with the first server, and wherein data packets associated with the VM are routed to the VM based upon the VM identifier and instruct a VM proxy running on a second server to receive the VM from the first server in response to the migration request, wherein the second server receives the VM, and wherein the VM name remains unchanged while the current VM address becomes associated with the second server after the second server receives the VM.