BGP Router Checkpointing for Fast Re-convergence
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current BGP routing protocols face inefficiencies in re-converging routing information after a restart, leading to prolonged downtime and excessive resource consumption, particularly in resource-constrained environments like military and mobile networks, where minimizing bandwidth and CPU usage is crucial.
Innovation Solution
Implementing fast synchronization capabilities by advertising network devices' ability to preserve forwarding state and establishing checkpoints during normal operation, allowing for rapid re-synchronization after a restart by requesting and sending only necessary update messages based on sequencing information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional BGP re-synchronization is performed after a restart, then routing information is exchanged between peers, but re-convergence time is prolonged and network performance degrades
Solution Approach 1:
The patent applies preliminary action by having the restarting router save its routing table state before the restart occurs. This saved state is then used after restart to quickly re-establish routing information without requiring a complete re-synchronization with all peers, thereby reducing re-convergence time while maintaining routing reliability.
Solution Approach 2:
The patent extracts only the necessary routing information that needs to be re-synchronized after a restart. By identifying and transmitting only the changed or updated routing entries rather than the entire routing table, the system reduces the time and resources required for re-convergence while ensuring routing information synchronization.
2Reliability
If complete routing information exchange is performed after restart, then routing tables are recreated, but computing resources and network bandwidth consumption increase
Solution Approach 1:
The patent extracts and transmits only the necessary routing information changes rather than the complete routing table. This selective synchronization reduces computing resource consumption and network bandwidth usage while still achieving reliable routing table synchronization across all peers.
Solution Approach 2:
The patent applies partial action by performing only the minimum necessary routing information exchange required for synchronization. Instead of transferring the entire routing table, the system identifies and transmits only the changed entries, reducing resource consumption while maintaining synchronization reliability.
3Reliability
If Graceful Restart is implemented to hold routes in stale state, then outage effect is reduced, but routing information still needs to be re-exchanged after restart
Solution Approach 1:
The patent combines with Graceful Restart by having the restarting router save its routing table state before restart and use this saved state after restart to quickly re-establish routing information. This preliminary action, when combined with GR's forwarding state preservation, reduces both the outage effect and the subsequent re-synchronization time.
Solution Approach 2:
The patent extracts only the necessary routing information changes for re-synchronization after a restart, building upon GR's mechanism. By identifying and transmitting only the changed routing entries rather than the entire table, the system reduces the time required for re-synchronization while maintaining the forwarding state preservation benefits of GR.
Data Source
AI summary
A checkpointing approach enables BGP peers to reduce the number of UPDATE messages that are exchanged and processed after a router restarts. A router receives update messages to update routing information for the one or more other devices, wherein each of the update messages includes a checkpoint marker and a route. The router stores the route in a checkpoint repository and stores the checkpoint markers in association with information identifying the other routers from which the checkpoint markers were received. In response to a restart event, the router sends each of the stored checkpoint markers to respective other devices and receives from the respective other devices only route updates that occurred later in time than the checkpoint markers.