IGP Proxy Signaling for Split SDN Controller Election
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In software-defined networks (SDNs), controller clusters face reliability issues due to connection failures between controllers, leading to multiple primary controllers being elected and causing network instability when the cluster splits into separate groups that lack knowledge of each other's status.
Innovation Solution
Implementing Interior Gateway Protocol (IGP) with new extensions to enable separated controllers or groups to communicate and share information about their status and connectivity using IGP proxies, such as IGP proxies with OSPF or IS-IS extensions, allowing correct election of a new primary controller even in split clusters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If controller clusters use traditional failure detection mechanisms, then simple implementation is achieved, but multiple primary controllers may be elected causing network instability
Solution Approach 1:
The patent introduces an Interior Gateway Protocol (IGP) as an intermediary mechanism between controllers in the cluster. The IGP enables indirect communication and status sharing between controllers through network infrastructure, allowing them to detect each other's operational status even when direct controller-to-controller communication fails. This mediator approach resolves the contradiction by providing reliable failure detection without requiring complex direct monitoring mechanisms between each controller pair.
Solution Approach 2:
The patent shifts the failure detection dimension from direct controller-to-controller communication to indirect detection through the network infrastructure using IGP. By changing the detection dimension to utilize network routing protocols and path analysis, controllers can determine each other's status through network path availability rather than direct connection status, thereby improving reliability without proportionally increasing complexity.
2Speed
If controllers maintain direct communication channels, then quick failure detection is achieved, but communication overhead increases
Solution Approach 1:
The patent makes the IGP communication infrastructure multi-functional by using it both for network routing purposes and for controller failure detection. The same network packets and protocols that route data traffic also carry controller status information and enable failure detection. This universal use of the communication infrastructure achieves fast detection without additional dedicated communication channels, thereby avoiding extra energy consumption and overhead.
3Adaptability or versatility
If controllers in split clusters operate independently, then autonomous operation is achieved, but incorrect primary controller election occurs
Solution Approach 1:
The patent implements feedback mechanisms through IGP where controllers continuously monitor network paths to other controllers and adjust their primary controller status based on this feedback. When a controller detects that the path to the current primary controller is unavailable through IGP routing information, it receives feedback indicating the primary controller may have failed, triggering a re-election process. This feedback loop ensures that even in split clusters, controllers can make informed decisions about primary controller election rather than operating blindly independently.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A method performed by a network controller for providing network reliability. The method includes employing an Interior Gateway Protocol (IGP) proxy configured to communicate controller information about the network controller to the one or more network elements using IGP (Link-State Advertisement (LSA) or Link State Protocol Data Unit (LSP)) with a Controllers TLV. The method uses the controller information for proper election of a primary controller in the event of a network failure.