Network Controller Traffic Rate Limiting for Data Center Latency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Data center networks face significant queuing delays and latency due to congestion, particularly triggered by excessive traffic, which negatively impacts overall network performance and can lead to dropped packets and buffer overflow conditions.
Innovation Solution
A scalable system that utilizes traffic matrix information, network traffic load, and congestion information to proactively adjust end-to-end traffic rate limits, reducing queuing delays while maintaining network utilization by identifying and ranking flows based on traffic volume and adjusting rate limits for both highly utilized and underutilized network node interfaces.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If traffic rate limits are increased to improve network utilization, then network resource utilization is improved, but queuing delays and latency increase due to congestion
Solution Approach 1:
The system dynamically adjusts traffic rate limits based on real-time network conditions. The network controller continuously monitors pause frame distribution and traffic matrix information, then adaptively modifies rate limits for different flows to balance utilization and latency requirements
Solution Approach 2:
Different traffic flows receive different rate limit adjustments based on their specific characteristics and network conditions. The system identifies individual flows causing congestion and applies targeted rate limit reductions only to those flows, while maintaining higher rates for non-congesting flows
2Loss of time
If traffic rate limits are reduced to decrease queuing delays, then queuing delays and latency are reduced, but network resource utilization decreases
Solution Approach 1:
The system applies rate limit reductions only partially - specifically to the extent necessary to eliminate congestion caused by particular flows. Rather than uniformly reducing all traffic, the system identifies and targets only the excess traffic contributing to queuing delays, maintaining optimal utilization for the remainder
3Loss of time
If proactive traffic rate limit adjustment is implemented to reduce congestion, then queuing delays are reduced, but system complexity increases due to additional monitoring and control mechanisms
Solution Approach 1:
The system uses pause frame distribution information as feedback to automatically detect congestion and trigger rate limit adjustments. The network controller monitors pause frames generated by network switches and uses this feedback to dynamically modify traffic rate limits, creating a self-regulating system that reduces manual configuration complexity
Data Source
AI summary
Methods and apparatus for determining and adjusting traffic rate limits to reduce queuing delays and latency in a data center network. In various embodiments, a network controller receives network topology information and traffic data relating to flows of the network. The traffic data includes traffic matrix information, flow path information, and pause frame distribution information. Based on this information, the network controller identifies server network node interfaces that are the source of a pause frame. Flows associated with such network node interfaces are also identified and ranked based on traffic volume. For ranked flows meeting predetermined criteria, reduced traffic rate limits are calculated and sent to associated server network node interfaces. In some embodiments, the reduced traffic rate limits are incrementally recovered when normal traffic resumes on a paused link. In further embodiments, traffic rate limits are also adjusted for flows of highly utilized and underutilized network node interfaces.