Network Latency Reduction via Foreground Traffic Prioritization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data bottlenecks in data centers lead to increased latency and dropped packets, negatively impacting user experience due to the filling of switch buffers by low-priority background traffic, which reduces available buffer space for high-priority foreground traffic.
Innovation Solution
Implementing an application control engine that identifies and prioritizes foreground traffic, using latency metrics like round-trip time to divert background flows to a delay buffer or delay acknowledgments for background traffic, thereby reducing latency and maintaining buffer space for priority traffic.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If background traffic flows are allowed to fill switch buffers, then buffer utilization is improved, but latency for foreground traffic increases and buffer space for priority traffic is reduced
Solution Approach 1:
The patent applies local quality by treating different traffic flows differently based on their priority. Background flows and foreground flows are subjected to different buffer management policies - background flows are delayed or dropped while foreground flows receive preferential treatment with guaranteed buffer space and lower latency, creating localized quality differences within the buffer management system
Solution Approach 2:
The patent implements asymmetry by creating unequal treatment of traffic flows. The buffer management system intentionally creates asymmetric behavior where background traffic experiences delays and restrictions while foreground traffic enjoys priority access, using asymmetric buffer allocation and flow control mechanisms
2Productivity
If background traffic is prioritized for buffer space, then buffer efficiency is improved, but user experience for priority applications deteriorates
Solution Approach 1:
The patent applies segmentation by dividing network traffic into distinct segments - background flows and foreground flows - with different priority levels. This segmentation allows the system to manage buffer resources differently for each segment, ensuring that priority applications receive adequate buffer space and performance while background traffic uses remaining resources
Solution Approach 2:
The patent introduces an intermediary mechanism (the buffer management system with flow classification and priority handling) that mediates between background traffic needs for buffer space and foreground traffic needs for low latency. This intermediary actively manages buffer allocation and flow control to balance efficiency and reliability
3Device complexity
If data bottlenecks are allowed to occur, then network simplicity is maintained, but packet loss increases and retransmissions are required
Solution Approach 1:
The patent applies preliminary action by proactively managing buffer space and traffic flows before bottlenecks occur. The system monitors buffer utilization and traffic patterns, and takes preventive actions such as delaying background flows or dropping low-priority packets before the buffer becomes completely full, thereby preventing bottlenecks and packet loss before they happen
Data Source
AI summary
A method is provided in one example embodiment that includes identifying a foreground flow that has a higher priority than at least one background traffic flow in a network. The method also includes determining (e.g., periodically) a latency metric associated with the foreground flow, and activating a latency reduction response if the latency metric exceeds a threshold. In more particular embodiments, the latency reduction response may include diverting background flows in the network to a buffer for a delay period and/or delaying acknowledgements for background flows until the latency metric is below the threshold. In yet more particular embodiments, the latency metric may be a round-trip time.