VM Migration Based on NUMA Device Proximity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current information handling systems face challenges in optimizing virtual machine performance and efficient resource utilization, particularly in virtualized environments, where virtual machines often experience overheads due to the need for virtual-physical device translation.
Innovation Solution
The system automatically selects a processor for a virtual machine based on proximity information, such as distance from physical devices, using a System Locality Information Table to associate the virtual machine with the nearest processor, thereby optimizing performance and reducing translation overheads.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If virtual machines use host operating system drivers to communicate with physical devices, then device compatibility and ease of operation are improved, but system performance and resource utilization efficiency deteriorate due to translation overheads
Solution Approach 1:
The patent extracts the driver functionality from the host operating system and places it directly in the virtual machine. This allows the virtual machine to communicate with physical devices using its own drivers, eliminating the need for virtual-physical translation through host drivers and thereby removing the associated performance overhead while maintaining device compatibility
Solution Approach 2:
The patent introduces a device mapping mechanism that directly maps physical devices to virtual machines based on proximity information. This intermediary mapping layer enables direct communication paths between virtual machines and physical devices, bypassing the traditional host driver translation layer and improving system performance
2Productivity
If virtual machines are allocated to processors without considering device proximity, then processor load balancing is improved, but communication efficiency with physical devices deteriorates
Solution Approach 1:
The patent applies local quality by considering the specific proximity relationship between each virtual machine, its associated physical device, and available processors. The system evaluates device proximity information to determine optimal processor assignments, ensuring that each virtual machine is allocated to the processor that provides the most efficient communication path to its required physical device
3Adaptability or versatility
If virtual machine migration is performed without considering device proximity, then system flexibility and adaptability are improved, but performance optimization deteriorates
Solution Approach 1:
The patent implements feedback mechanisms that continuously monitor device proximity information and virtual machine performance metrics. Based on this feedback, the system dynamically adjusts virtual machine processor allocations and migration decisions to maintain optimal performance while preserving system flexibility and adaptability
Data Source
AI summary
In one embodiment, a system includes a device and first and second processors. The first and second processors are configured to access first information. The first information includes a first parameter and a second parameter. The first parameter is associated with a position of the device relative to the first processor. The second parameter is associated with a position of the device relative to the second processor. The system is configured to access second information associated with a virtual machine. The virtual machine is configured to run on a host operating system. The virtual machine includes a guest operating system configured to communicate with the device. The system is also configured to automatically select the first processor based at least on the first information and the second information and to associate the virtual machine with the first processor in response to automatically selecting the first processor.