Stackable Router Synchronization via Master-Slave Protocol Updates
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current routing protocols, such as OSPF, do not provide a means to synchronize routing information across multiple physical routers in a stack, leading to inefficiencies in load distribution and resilience, particularly when a master router fails, as back-up routers are only activated upon failure and do not contribute to traffic routing until then.
Innovation Solution
Implementing a synchronization mechanism where a master router sends direct update messages to slave routers, allowing them to operate in sync with the master's protocol state machinery and link state database, enabling seamless transition and load distribution without the need for a dedicated back-up unit, using TCP for reliable data transfer and verification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If a master-router backup system is implemented where only the master router performs full routing protocol operations, then device complexity is reduced and manufacturing efficiency is improved, but reliability deteriorates because back-up routers cannot take over routing duties immediately upon master failure
Solution Approach 1:
The patent applies preliminary action by having slave routers pre-synchronize their link state databases with the master router before failure occurs. The master router periodically sends database descriptions to slave routers, allowing them to maintain up-to-date routing information in advance. This ensures that when master failure occurs, slave routers can immediately assume routing responsibilities without needing to perform discovery or synchronization procedures, thus resolving the contradiction between reduced complexity and maintained reliability.
2Reliability
If slave routers fully synchronize their link state databases with the master router, then reliability and seamless transition capability are improved, but bandwidth consumption increases due to continuous data transfer requirements
Solution Approach 1:
The patent extracts only the essential routing information needed for synchronization by using database description packets that contain summaries of link state changes rather than transferring complete database contents. The master router sends only the necessary updates to slave routers, and slave routers selectively synchronize only the changed portions of the link state database. This extraction approach maintains reliability through comprehensive synchronization while minimizing bandwidth consumption by avoiding redundant data transfer.
3Productivity
If routing information is synchronized continuously across all slave routers, then load distribution capability is improved, but device complexity increases due to synchronization mechanism requirements
Solution Approach 1:
The patent implements feedback mechanisms where slave routers send database description packets to the master router, indicating which link state entries have changed or need updating. The master router uses this feedback information to determine what data needs to be transmitted during synchronization. This feedback-based approach enables efficient load distribution by keeping all slave routers informed of network changes while reducing the complexity of synchronization mechanisms through targeted, incremental updates rather than comprehensive synchronized transfers.
Data Source
AI summary
A stack of network routers is composed of at least one (elected) master unit and one or more slave units each capable of running a routing protocol. Only the master unit runs the entire routing protocol at a given time. It forwards direct update messages via a transmission control protocol to each of the slave units. These direct update messages may include specific packet formats for the protocol state machinery where such machinery is required by the routing protocol, e.g. the interface state machine and the neighbor state machine for the OSPF protocol, and for the net databases, e.g. the link state databases for the OSPF protocol, in each of the slave units. Each slave unit may run its protocol state machinery (where provided) based purely on the direct update messages received from the master. The synchronization of the net databases may be based on snooping net update packets and a comparison of the information received thereby with verification data messages sent from the master unit. The synchronization allows the running of the routing protocol on multiple physical routers in a stack without requiring the overhead of a back-up unit that would not perform routing unless a master router unit became unavailable.