Virtual Switch Congestion Notification for VM Queue Isolation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In virtual networking environments, conventional flow control methods fail to efficiently manage congestion as they cannot distinguish between individual virtual machine queues, leading to unnecessary throttling of all queues when congestion occurs in one, resulting in decreased efficiency.
Innovation Solution
A computer program product and method that determines congestion in a virtual switching device by identifying the source of excess data packets and sends a notification to the source to decrease its transmission rate, allowing for individualized congestion control rather than throttling all associated virtual sources.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional flow control is used in virtual networking, then reliability is improved by preventing buffer overflows, but productivity deteriorates because all virtual machine queues are throttled when congestion occurs in any single queue
Solution Approach 1:
The patent segments the flow control mechanism by introducing virtual switch instances that operate at the virtual machine level, allowing independent congestion management for each virtual machine queue rather than applying flow control to the entire physical network interface card. This segmentation enables selective throttling of only congested queues while maintaining throughput for non-congested queues.
Solution Approach 2:
The patent applies local quality by implementing congestion control at the specific location where congestion occurs (individual virtual machine queues) rather than applying uniform control across all queues. Each virtual switch instance monitors and controls its associated virtual machine's queue independently, allowing non-congested queues to maintain high transmission rates while only congested queues receive flow control signals.
2Reliability
If flow control blocks incoming traffic to backed up queues, then reliability is improved by preventing data loss, but loss of substance increases because network resources are wasted transmitting blocked frames
Solution Approach 1:
The patent implements preliminary action by having virtual switch instances proactively monitor queue buffer levels and send flow control signals to sources before buffers become completely full. This early intervention prevents buffer overflows and data loss while avoiding the waste of transmitting frames that would be immediately dropped, as the flow control signal is sent in advance to slow down the source transmission rate.
3Reliability
If all queues receive block commands during congestion, then reliability is improved by preventing buffer overflows, but device complexity increases due to the need for granular queue management
Solution Approach 1:
The patent introduces virtual switch instances as intermediary components between the physical network interface card and virtual machine queues. These intermediaries handle the complex task of monitoring individual queue states and generating targeted flow control signals, simplifying the overall architecture by centralizing granular control logic in the virtual switch layer rather than requiring complex modifications to the physical NIC or hypervisor.
Data Source
AI summary
Congestion management for data traffic in a virtual domain identifies a congestion source and sends a message to the source to adjust data traffic rates. The source may be a virtual machine hosted by a physical server with one or more virtual servers incorporated. A congestion manager may identify the source and send the message to the source without affecting other data sources hosted by the physical server or the virtual servers. In some embodiments, information about the congestion source may be encapsulated in a packet payload readable only by the congestion source so only the congestion source receives the instruction to adjust the transmission rate.