Cisco SSO vs NSF Operations

Stateful Switchover (SSO) Operation

SSO makes one of the supervisor engines active while the other is on standby. SSO then synchronizes data between the two supervisor engines. When the current supervisor engine fails, is removed from the switch, or is manually turned down for maintenance, a switchover from the active to the redundant supervisor engine occurs. This form of switchover is guaranteed not to disrupt Layer 2 traffic.

SSO preserves state information between the two supervisor engines after initial synchronization, including forwarding data. At startup and anytime the configuration of the active supervisor engine is modified, configuration data and data structures are synced from the active to the redundant supervisor engine. To ensure that the redundant supervisor engine is always prepared to take over in the event of an issue with the active supervisor engine, both supervisor engines in networking devices that support SSO must be operating in the same configuration. SSO switchover can also continue to route Layer 3 traffic after a switchover and keeps FIB and adjacency entries.

Switchover occurs when the redundant supervisor engine takes over for the active supervisor engine and executes the routing protocol. The switchover time from the active to the redundant supervisor engine is 0 to 3 seconds.

Nonstop Forwarding (NSF) Operation

After a supervisor engine switchover, NSF’s primary goal is to keep forwarding IP traffic. To reduce the amount of time a network is inaccessible to its users after a switchover, NSF collaborates with SSO. SSO is always used by Cisco NSF, which offers redundancy for Layer 3 communications.

BGP, OSPF, EIGRP, and IS-IS are routing protocols that support Cisco NSF, while Cisco Express Forwarding (CEF) is a forwarding protocol that supports Cisco NSF. The addition of NSF capability and awareness to the routing protocols allows routers to use them to detect a switchover and take the appropriate steps to keep forwarding network traffic and get route information from the peer devices. In place of information obtained from peer devices, the IS-IS protocol can be configured to recover route information after a switchover using state information synchronized between the active and redundant supervisor engines.

If networking equipment is using NSF-compliant software, it is NSF-aware. If a device has been set up to support NSF, it is NSF-capable and will reassemble its routing data from neighbors who are either NSF-aware or NSF-capable.

During switchover, each protocol relies on CEF to maintain packet forwarding while the routing protocols reconstruct the Routing Information Base (RIB) tables. Following convergence of the routing protocols, CEF modifies the FIB table and deletes outdated route information. The new FIB data is subsequently updated in the modules via CEF.