I/O Virtualization Aggregation Device for VM Network Offload
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtualized data center environments, the overhead associated with input/output (I/O) virtualization operations, such as high latency and inefficient CPU utilization, is caused by processor-intensive 'world switching' performed by the hypervisor, which degrades VM performance and bandwidth utilization.
Innovation Solution
An aggregation system with a virtual switch at a server node aggregation device offloads virtual switching functions from the hypervisor, establishing a direct communication path between virtual machines and physical devices, reducing the hypervisor's role in I/O operations and eliminating CPU-intensive processing.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If the hypervisor performs processor-intensive 'world switching' for I/O virtualization operations, then virtualization functionality is provided, but I/O latency increases and CPU utilization becomes inefficient
Solution Approach 1:
The patent extracts the I/O virtualization functions from the hypervisor and places them in a dedicated I/O virtualization device. This separation removes the time-consuming world switching operations from the CPU path, allowing I/O operations to proceed without the latency introduced by hypervisor-mediated context switching between physical and virtual addresses.
Solution Approach 2:
The patent introduces an I/O virtualization device as an intermediary component between the virtual machines and physical I/O devices. This mediator handles address translation and I/O operations independently, eliminating the need for the hypervisor to intervene in each I/O operation and thereby reducing I/O latency.
2Adaptability or versatility
If the hypervisor performs processor-intensive 'world switching' for I/O virtualization operations, then virtualization functionality is provided, but CPU resource consumption increases
Solution Approach 1:
The patent extracts I/O virtualization operations from the hypervisor's responsibility, placing them in a dedicated I/O virtualization device. This extraction eliminates the CPU-intensive world switching operations from the hypervisor's workload, significantly reducing CPU resource consumption while maintaining full virtualization functionality.
Solution Approach 2:
The I/O virtualization device performs self-service by autonomously handling address translation and I/O operations without requiring hypervisor intervention. The device maintains its own translation lookaside buffers and performs I/O operations directly, freeing CPU resources while preserving virtualization capabilities.
3Adaptability or versatility
If virtual machines share physical devices through hypervisor-mediated virtualization, then resource consolidation is achieved, but I/O performance and bandwidth utilization degrade
Solution Approach 1:
The patent introduces an I/O virtualization device as an intermediary that enables multiple virtual machines to share physical I/O devices without the performance degradation caused by hypervisor-mediated virtualization. This mediator provides direct I/O access paths and maintains translation buffers, preserving bandwidth utilization while enabling resource consolidation.
Solution Approach 2:
The patent segments the virtualization functionality into separate components: the hypervisor maintains resource allocation and scheduling, while the I/O virtualization device handles I/O operations and address translation. This segmentation allows resource consolidation to be maintained while I/O performance is preserved through dedicated hardware handling.
Data Source
AI summary
Described is an aggregation device comprising a plurality of virtual network interface cards (vNICs) and an input/output (I/O) processing complex. The vNICs are in communication with a plurality of processing devices. Each processing device has at least one virtual machine (VM). The I/O processing complex is between the vNICs and at least one physical NIC. The I/O processing complex includes at least one proxy NIC and a virtual switch. The virtual switch exchanges data with a processing device of the plurality of processing devices via a communication path established by a vNIC of the plurality of vNICs between the at least one VM and at least one proxy NIC.