Map Server Cache Update for Virtual Machine Migration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing changes in virtualized networks with overlays in large-scale implementations is challenging due to limitations in propagating changes, leading to latency and scalability issues, particularly during virtual machine migrations.
Innovation Solution
A proactive method for updating distributed caches in network devices by using a map server to notify subscribed edge devices of virtual machine changes, ensuring that cached mapping data remains current and reducing unnecessary packet transmissions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If distributed caches are updated reactively in response to virtual machine changes, then network devices can maintain cached mapping data, but latency increases and cache information becomes outdated during virtual machine migrations
Solution Approach 1:
The map server proactively identifies and notifies network devices about upcoming virtual machine migrations before they occur. By performing the notification action in advance, the system ensures that network devices can update their caches timely, preventing stale data from being used and reducing latency during actual migrations.
2Reliability
If map servers notify all network devices of every virtual machine change, then cache information remains current, but network traffic and processing overhead increase
Solution Approach 1:
The system implements selective notification where the map server identifies and notifies only those network devices that have cached mapping data for the specific virtual machine being migrated. This localized approach ensures that each network device receives notifications only when relevant to its cached data, reducing unnecessary network traffic while maintaining cache freshness where needed.
3Productivity
If network devices maintain distributed caches of mapping data, then packet forwarding efficiency improves, but scalability is limited by the ability to propagate changes across the network
Solution Approach 1:
The map server implements a feedback mechanism where it tracks which network devices have cached mapping data for virtual machines. When a virtual machine migration is detected or anticipated, the map server uses this feedback information to selectively notify the appropriate network devices. This feedback-driven approach enables the system to scale efficiently by maintaining cache consistency only where necessary, rather than requiring broad network-wide propagation of all changes.
Data Source
AI summary
Technologies are provided in example embodiments for associating a subscriber list to mapping data of a virtual machine, adding subscriber information of a network device to the subscriber list when a map request for the mapping data is received from the network device, and purging the subscriber information from the subscriber list when a preconfigured time period assigned to the subscriber information expires. In particular embodiments, the subscriber information includes an identification of the network device and the mapping data includes a virtual address of the virtual machine mapped to a physical address of the virtual machine. More specific embodiments include sending a notification signal with new mapping data of the virtual machine to each one of one or more network devices identified in corresponding subscriber information stored in the subscriber list. In further specific embodiments, the network device is either a map server or a map resolver.