Virtual Network Congestion Notification via Layered Feedback
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current network systems face inefficiencies and increased congestion due to the long round trip time for packet transmission and acknowledgement, leading to delayed recognition of congestion, especially when it occurs within the same host machine, resulting in unnecessary packet retransmissions and worsened congestion.
Innovation Solution
A congestion notification system that sends messages from lower layers to higher layers in a computer virtualization networking stack, allowing for early reduction of packet transmission rates to alleviate congestion, resuming full speed only when congestion is cleared or notification of reduced congestion is received.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the source system waits for acknowledgement messages before determining congestion, then packet transmission reliability is maintained, but congestion detection is delayed causing unnecessary retransmissions
Solution Approach 1:
The patent implements preliminary congestion detection by monitoring queue depths at intermediate network nodes before packets are lost. The network device detects congestion when the queue depth exceeds a threshold and sends a congestion notification message to the source system, enabling early congestion detection without waiting for packet loss and retransmission cycles to occur.
Solution Approach 2:
The patent establishes a feedback mechanism where intermediate network devices continuously monitor their queue depths and provide real-time congestion status information back to the source system through congestion notification messages. This feedback loop enables the source system to adjust its transmission rate dynamically based on actual network conditions rather than waiting for acknowledgment-based detection.
2Productivity
If the source system sends multiple packets while waiting for acknowledgement, then transmission efficiency is improved, but congestion is worsened due to increased packet volume
Solution Approach 1:
The patent implements dynamic transmission rate adjustment where the source system continuously adapts its packet sending rate based on real-time congestion notifications from the network. When congestion is detected, the source system reduces its transmission rate; when congestion clears, it increases the rate again. This dynamic adjustment optimizes transmission efficiency while preventing congestion exacerbation.
Solution Approach 2:
The feedback mechanism allows the source system to receive real-time congestion status information and adjust its transmission behavior accordingly. The source system uses the congestion notification messages to modulate its packet injection rate into the network, achieving a balance between maintaining high transmission efficiency and avoiding congestion.
3Productivity
If the round trip time is long for geographically distant systems, then more packets can be sent in parallel, but congestion recognition is significantly delayed
Solution Approach 1:
The patent places congestion detection functionality at intermediate network nodes along the transmission path, allowing congestion to be detected locally and immediately reported back to the source system. This eliminates the dependency on long round-trip acknowledgment messages for congestion detection, enabling early congestion recognition even in geographically distributed networks with long propagation delays.
Solution Approach 2:
The patent introduces intermediate network devices as mediators that perform local congestion monitoring and send direct congestion notification messages to the source system. These intermediary nodes act as early warning sensors that detect congestion conditions before they propagate through the entire network path, providing timely feedback to the source system independent of the destination system's location.
4Object-generated harmful factors
If the source system reduces packet transmission rate upon congestion detection, then congestion is alleviated, but transmission productivity decreases
Solution Approach 1:
The patent implements dynamic rate adjustment where the source system continuously adapts its transmission rate based on real-time congestion notifications. Rather than maintaining a permanently reduced rate, the system dynamically increases the rate when congestion clears and decreases it when congestion occurs, optimizing the balance between congestion control and transmission productivity over time.
Solution Approach 2:
The feedback mechanism enables the source system to receive continuous congestion status information and adjust its transmission rate accordingly. The source system uses this feedback to modulate its packet injection rate, reducing transmission when congestion is present to alleviate network load, and increasing transmission when the network is clear to maintain high productivity.
Data Source
AI summary
The congestion notification system of some embodiments sends congestion notification messages from lower layer (e.g., closer to a network) components to higher layer (e.g., closer to a packet sender) components. When the higher layer components receive the congestion notification messages, the higher layer components reduce the sending rate of packets (in some cases the rate is reduced to zero) to allow the lower layer components to lower congestion (i.e., create more space in their queues by sending more data packets along the series of components). In some embodiments, the higher layer components resume full speed sending of packets after a threshold time elapses without further notification of congestion. In other embodiments, the higher layer components resume full speed sending of packets after receiving a message indicating reduced congestion in the lower layers.