Network Function Virtualization Queue Congestion Control
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In network function virtualization (NFV), efficient sharing of specialized hardware like GPUs among multiple network functions in a network function chain (NFC) is challenging, leading to potential bottlenecks and reduced throughput.
Innovation Solution
A computer system with a scheduler that monitors receive queue occupancy and adjusts packet flow rates between network functions using a congestion monitor and rate controller, ensuring efficient resource allocation and preventing bottlenecks.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple network functions share a single GPU in an NFC, then resource utilization and cost-effectiveness improve, but performance bottlenecks and packet loss increase
Solution Approach 1:
The patent implements dynamic rate adjustment mechanisms that continuously monitor queue occupancy and adapt packet flow rates in real-time. The rate controller dynamically modifies the rate of packet flow between upstream NFs and the shared GPU based on current load conditions, allowing the system to optimize performance under varying traffic conditions while preventing bottlenecks
Solution Approach 2:
The patent employs feedback loops where the congestion monitor continuously reports queue occupancy status to the rate controller, which then adjusts packet flow rates accordingly. This closed-loop control system uses occupancy feedback to prevent packet loss and maintain optimal throughput when multiple NFs share a GPU resource
2Productivity
If packet flow rate is increased to improve throughput, then processing speed improves, but queue overflow and packet loss increase
Solution Approach 1:
The patent implements preliminary rate limiting actions before queue overflow occurs. The congestion monitor detects rising queue occupancy levels and triggers rate adjustments in advance, preventing packet loss by proactively controlling the inflow rate to match the processing capacity of downstream NFs
Solution Approach 2:
The patent dynamically changes the packet flow rate parameter based on queue occupancy conditions. The rate controller adjusts this parameter in real-time, modifying it from high to low rates depending on whether the queue is under or over capacity, thereby maintaining both high throughput and packet delivery reliability
3Reliability
If a single NF独占 a GPU to ensure high performance, then processing reliability improves, but resource utilization and scalability deteriorate
Solution Approach 1:
The patent enables a single GPU to serve multiple network functions simultaneously through the shared NFC architecture. The scheduler and rate controller coordinate multiple NFs to share the GPU resource, allowing one device to perform multiple functions while maintaining performance through dynamic resource allocation and flow rate control
Data Source
AI summary
An example computer system includes a hardware platform including a processing unit and software executing on the hardware platform. The software includes a workload and a scheduler, the workload including a network function chain having network functions, the scheduler configured to schedule the network functions for execution on the processing unit. A downstream network function includes a congestion monitor configured to monitor a first receive queue supplying packets to the downstream network function, the congestion monitor configured to compare occupancy of the first receive queue against a queue threshold. An upstream network function including a rate controller configured to receive a notification from the congestion monitor generated in response to the occupancy of the first receive queue exceeding the queue threshold, the rate controller configured to modify a rate of packet flow between a second receive queue and the upstream network function in response to the notification.