System and method for graceful restart

Abstract

A system for maintaining routing capabilities in a router having a failed control plane provides an active control plane in the router in communication with at least one external node, the active control plane running at least one routing process. A backup control plane may be interconnected with the active control plane, so that the active control plane may periodically transmit synchronization signals to the backup control plane. The backup control plane may update its state based on these synchronization signals. Moreover, the backup control plane may be programmed to take over the routing process of the active control plane if the active control plane fails.

1 . A router, comprising:
 a primary control plane running one or more routing processes;   a plurality of secondary control planes interconnected with the primary control plane in a ring topology, wherein each control plane is connected to a downstream neighbor in the ring;   wherein:
 the primary control plane periodically transmits synchronization signals indicating a forwarding state of the primary control plane to the secondary control planes; 
 the secondary control planes update their states based on the synchronization signals; 
 the synchronization messages are transmitted individually through each control plane in the ring; and 
 in the event that the primary control plane fails, its next available downstream neighbor in the ring takes over its routing processes. 
   
     
     
         2 . The router according to  claim 1 , wherein the synchronization signals are transmitted with a corresponding sequence number; and
 wherein each secondary control plane is capable of detecting that it did not receive a signal based on the sequence number of a received signal.   
     
     
         3 . The router according to  claim 2 , wherein a given secondary control plane detects that it did not receive a synchronization signal by detecting a gap in the sequence numbers of two non-consecutive synchronization messages. 
     
     
         4 . A router, comprising:
 an active control plane running one or more routing processes;   a plurality of inactive control planes interconnected with the active control plane;   wherein:
 the active control plane is serially connected to each inactive control plane; 
 the active control plane periodically transmits synchronization signals indicating a forwarding state of the active control plane to the inactive control plane; 
 the inactive control planes update their state based on the synchronization signals; 
 the synchronization messages are simultaneously transmitted to each inactive control plane; and 
 in the event that the active control plane fails, a first one of the plurality of inactive control planes initiates an election process to elect a inactive control plane to take over the routing processes of the failed active control plane. 
   
     
     
         5 . The router according to  claim 4 , wherein the election process is initiated by a first inactive control plane to detect the failure. 
     
     
         6 . The router according to  claim 5 , wherein, as part of the election process, the first inactive control plane to detect the failure nominates itself as the new active plane, and requests approval from the other inactive planes. 
     
     
         7 . The router according to  claim 5 , wherein the active control plane and the inactive control plane are arranged in a ring topology. 
     
     
         8 . A router, comprising:
 a primary control plane running one or more routing processes;   a secondary control plane interconnected with the primary control plane;   wherein the primary control plane periodically transmits synchronization signals indicating a forwarding state of the primary control plane to the secondary control plane, and the secondary control plane updates its state based on the synchronization signals;   wherein the synchronization signals are transmitted with a corresponding sequence number; and   wherein the secondary control plane is capable of detecting that it did not receive a signal based on the sequence number of a received signal.   
     
     
         9 . The router according to  claim 8 , wherein the secondary control plane detects that it did not receive a synchronization signal by determining that no synchronization signals were received within a predetermined time period. 
     
     
         10 . The router according to  claim 8 , wherein the secondary control plane detects that it did not receive a synchronization signal by detecting a gap in the sequence numbers of two non-consecutive synchronization messages. 
     
     
         11 . The router according to  claim 8 , wherein the secondary control plane is capable of detecting failure of the primary control plane. 
     
     
         12 . The router according to  claim 8 , wherein the secondary control plane is capable of taking over the one or more routing processes of the primary control plane upon failure of the primary control plane. 
     
     
         13 . The router according to  claim 12 , wherein taking over the one or more routing processes includes the secondary control plane establishing a user datagram protocol (“UDP”) with remote peers. 
     
     
         14 . The router according to  claim 8 , wherein the one or more routing processes implement a Border Gateway Protocol (BGP). 
     
     
         15 . The router according to  claim 8 , wherein each of the control planes has a preassigned Transmission Control Protocol (TCP) port number used to establish communication with remote peers. 
     
     
         16 . A method for managing routing connections in a router having an active control plane in communication with at least one external node, and a plurality of backup control planes, the active control plane and the plurality of backup control planes arranged in a ring topology, the method comprising:
 periodically transmitting synchronization signals from the active control plane to the plurality of backup control planes, wherein the synchronization messages are individually transmitted downstream from the active control plane through each of the backup control planes in the ring;   detecting a failure of the active control plane;   electing a first one of the backup control planes to serve as a new active control plane, the electing comprising selecting a next available downstream backup plane;   establishing communication between the new active control plane and the at least one external node; and   synchronizing the new active control plane with the at least one external node.   
     
     
         17 . A system for maintaining routing capabilities in a router having a failed control plane, comprising:
 an active control plane in the router in communication with at least one external node in a network, the active control plane running at least one routing process;   a backup control plane interconnected with the active control plane;   wherein the active control plane periodically transmits synchronization signals to the backup control plane, and the backup control plane updates its state based on the synchronization signals;   wherein the backup control plane is programmed to take over the routing process of the active control plane if the active control plane fails;   wherein the at least one node external to the router includes a peer router in the network; and   wherein the backup control plane establishes a connection with the peer router in the network in response to determining that the active control plane has failed.   
     
     
         18 . The system according to  claim 17 , wherein the backup control plane establishes the connection with the peer router in the network by transmitting an open message from the backup control plane to the peer router. 
     
     
         19 . A system for maintaining routing capabilities in a router having a failed control plane, comprising:
 an active control plane in the router in communication with at least one external node, the active control plane running at least one routing process;   a plurality of backup control planes interconnected with the active control plane, each backup control plane programmed to determine which backup control plane should take over in the event of a failure of the active control plane;   wherein the active control plane periodically transmits synchronization signals to the backup control planes, and the backup control planes updates their states based on the synchronization signals;   and wherein each backup control plane is programmed to take over the routing process of the active control plane if the active control plane fails.   
     
     
         20 . The router according to  claim 19 , wherein the synchronization signals are transmitted with a corresponding sequence number; and
 wherein each backup control plane is capable of detecting that it did not receive a signal based on the sequence number of a received signal.   
     
     
         21 . The router according to  claim 20 , wherein a given backup control plane detects that it did not receive a synchronization signal by detecting a gap in the sequence numbers of two non-consecutive synchronization messages.