Backup Router Multicast Recovery for Network Convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In multi-access network segments, the current method of relying on a designated router (DR) and backup designated router (BDR) for routing updates is insufficient for real-time applications like voice and video conferencing, as it leads to delays in DR failure recovery due to the need for extensive unicast messages, which waste network resources and cause network convergence issues.
Innovation Solution
Implementing a method where the BDR sends a multicast update immediately upon detecting a silent DR, and sets a non-zero re-flood time to expedite the recovery process, reducing the need for numerous unicast messages and minimizing network resource consumption.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the BDR sends unicast messages to each unresponsive node after DR failure, then reliability of routing update delivery is improved, but network resource consumption and recovery time increase
Solution Approach 1:
The BDR pre-positions itself as a standby router with all necessary routing information before the DR fails. When failure occurs, the BDR immediately activates and sends multicast updates to all nodes, eliminating the delay of gradual unicast delivery. This preliminary preparation enables instant failover while maintaining reliable update delivery to all nodes.
2Reliability
If the BDR sends unicast messages to each unresponsive node, then complete delivery of routing updates is ensured, but network bandwidth and processing capacity are consumed
Solution Approach 1:
Instead of sending separate unicast messages to each unresponsive node, the BDR merges all routing updates into a single multicast message that is simultaneously delivered to all nodes on the network segment. This consolidation reduces network bandwidth consumption and processing overhead while ensuring complete delivery of routing information to all receivers.
3Stability of the object's composition
If the re-transmission time is set to default values (5 seconds), then network stability is maintained, but real-time applications like voice and video conferencing experience delays
Solution Approach 1:
The system dynamically adjusts the re-transmission time parameter based on network conditions and application requirements. For real-time applications requiring fast convergence, the re-transmission time is reduced from the default 5 seconds to a shorter interval, enabling quicker DR failure detection and BDR activation while maintaining network stability through controlled update pacing.
Data Source
AI summary
In one embodiment, a method includes setting a re-flood time to a non-zero first value. Update data that indicates a change to link data is received over a network interface from an updating node of multiple nodes on a multi-access segment connected to the interface. The change is to link data that indicates links on the updating node. It is determined whether an acknowledgment message is not received within the re-flood time from a particular node of the multiple nodes. If an acknowledgment message is not received within the re-flood time, then a multicast packet directed to the multiple nodes and including the change to the link data is sent over onto the segment. This method enables rapid response to failure of a primary node designated to ensure reliable flooding of link data.