Switch Management via Distributed Controller Metrics
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Managing a large network of interconnected switches is inefficient due to bandwidth constraints and the complexity of configuring each switch individually, leading to potential bottlenecks and single points of failure.
Innovation Solution
A system of controllers that remotely discover and manage switches using multicast messages, distributing the management responsibility among multiple controllers based on metrics such as hop count, bandwidth, and load, allowing for scalable and high-availability network management.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple switches are interconnected to support large number of users, then network capacity increases, but management complexity and configuration difficulty increase
Solution Approach 1:
A controller is introduced as an intermediary device that manages switches remotely. The controller receives configuration instructions and translates them into switch-specific commands, eliminating the need for direct manual configuration of each switch and reducing management complexity as network size grows
Solution Approach 2:
The network management function is segmented into two parts: the controller that handles high-level management and configuration, and the switches that execute specific forwarding operations. This segmentation allows centralized management of large networks while keeping individual switch operations simple
2Loss of information
If discovery messages are flooded via all local interfaces to discover topology, then topology discovery is achieved, but bandwidth consumption increases
Solution Approach 1:
The controller acts as an intermediary that receives topology discovery responses from switches and consolidates topology information centrally. Switches send discovery responses to the controller instead of flooding responses across the entire network, significantly reducing bandwidth consumption while maintaining complete topology visibility
Solution Approach 2:
Multiple topology discovery responses from different switches are merged and consolidated by the controller into a unified topology view. This combining approach eliminates redundant traffic and allows efficient topology management without requiring each switch to flood messages independently
3Loss of information
If unicast-based message exchanges are used to share information between switches, then information sharing is achieved, but control traffic bandwidth consumption increases
Solution Approach 1:
The controller serves as a central intermediary for all information sharing between switches. Instead of switches exchanging unicast messages directly (which consumes bandwidth), all information is routed through the controller, which consolidates and manages communications efficiently, reducing overall control traffic bandwidth consumption
4Ease of operation
If a single controller manages all switches, then centralized management is achieved, but single point of failure risk increases
Solution Approach 1:
The centralized management function is segmented across multiple controllers instead of being concentrated in a single device. Each controller manages a subset of switches, providing centralized management capabilities while distributing the failure risk. If one controller fails, others continue to manage their respective switches independently
Solution Approach 2:
The system changes the parameter of controller quantity from one to multiple, transforming the architecture from a single-point centralized management to a distributed centralized management model. This parameter change maintains the ease of operation benefits of centralized management while eliminating the single point of failure vulnerability
Data Source
AI summary
One embodiment of the present invention provides a computing system. The computing system includes a discovery module, a high-availability management module, and a controlling module. The discovery module determines local switch-specific information associated with a switch based on a discovery response packet. The high-availability management module determines remote switch-specific information about the same switch with respect to a remote computing system. The controlling module determines whether the computing system is to manage the switch based on a metric derived from the local and remote switch-specific information.