Distributed SDN VNF Monitoring via Data Plane Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In software-defined networking (SDN) environments, centralized application layer monitoring consumes significant bandwidth and compute resources due to the need for constant communication between the SDN controller and data plane nodes for virtual network function (VNF) monitoring, leading to scalability and latency issues.
Innovation Solution
Distributing monitoring tasks from the SDN controller to data plane nodes (DPNs), where DPNs generate and consume monitoring request and response messages independently, using flow table entries to match byte sequences and identify proper VNF operation, thereby reducing the controller's workload.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If centralized monitoring is implemented at the SDN controller, then monitoring coverage is comprehensive, but bandwidth and compute resources are consumed excessively
Solution Approach 1:
The patent segments the centralized monitoring function into distributed monitoring capabilities at the data plane node level. Each data plane node independently performs monitoring of its local VNFs, dividing the monolithic monitoring task into smaller, manageable units that can be executed locally, thereby reducing the bandwidth and compute resource burden on the SDN controller while maintaining comprehensive monitoring coverage
Solution Approach 2:
Data plane nodes are empowered to perform self-monitoring of their local VNFs without requiring constant controller intervention. The nodes autonomously generate monitoring requests, process responses, and detect failures locally, enabling the system to serve its own monitoring needs without excessive reliance on centralized controller resources
2Measurement precision
If centralized monitoring processes all VNF monitoring traffic, then monitoring accuracy is maintained, but latency increases due to controller bottleneck
Solution Approach 1:
By segmenting monitoring operations to occur locally at data plane nodes rather than centrally at the controller, the patent eliminates the bottleneck effect. Local monitoring operations complete faster since they don't require round-trip communication with the controller for each monitoring event, thereby reducing latency while maintaining accuracy through proper configuration and periodic controller verification
Solution Approach 2:
The system performs preliminary configuration of monitoring parameters, byte sequences, and flow table entries at the data plane nodes before actual monitoring begins. This preliminary setup enables nodes to perform accurate monitoring locally without requiring real-time controller involvement, thus maintaining monitoring accuracy while reducing operational latency
3Productivity
If monitoring is distributed to data plane nodes, then resource efficiency improves, but system complexity increases
Solution Approach 1:
The patent implements a universal monitoring framework where data plane nodes perform multiple functions: they act as both network forwarding elements and local monitoring agents. The same data plane node infrastructure handles both packet forwarding and VNF monitoring, eliminating the need for separate monitoring hardware or software and thereby managing complexity while improving resource efficiency
Solution Approach 2:
The patent introduces flow table entries as intermediaries that mediate between the distributed monitoring operations at data plane nodes and the centralized controller. These flow table entries encode monitoring rules, byte sequences, and response handling logic, serving as a intermediary mechanism that simplifies the interaction between distributed nodes and the controller, thereby managing system complexity
4Adaptability or versatility
If all monitoring responses pass through the controller, then centralized control is maintained, but scalability is limited
Solution Approach 1:
The patent segments the monitoring response handling into two paths: critical responses that require controller awareness and routine responses that can be handled locally. This segmentation allows the system to scale by enabling data plane nodes to independently process and filter monitoring responses, reducing the burden on the controller and enabling the system to handle larger numbers of VNFs without proportionally increasing controller load
Solution Approach 2:
The controller performs partial monitoring control by setting up flow table entries and monitoring rules at data plane nodes, rather than processing every monitoring response. This partial action approach allows the system to scale efficiently, as the controller only needs to maintain monitoring state for a fraction of the total monitored VNFs, while data plane nodes handle the bulk of monitoring operations autonomously
Data Source
AI summary
A method is implemented by a network device to reduce bandwidth and compute resources needed by the controller of a software defined networking (SDN) controller by distributing monitoring of virtual network functions (VNFs) to data plane nodes (DPNs) in a SDN network. The method includes receiving a monitoring request from the SDN controller to monitor a VNF, receiving a byte sequence from the SDN controller for the VNF, and generating flow table entries for monitoring the VNF by matching the byte sequence with monitoring response messages from the VNF.