On a Catalyst chassis with two supervisor modules, which feature lets you remove the active supervisor without disrupting forwarding?

Stateful Switchover (SSO) is the correct answer because it allows a redundant supervisor engine to take over seamlessly if the active supervisor fails or is removed, without interrupting data forwarding. SSO synchronizes the state information (like MAC address tables, routing tables, and CEF entries) between the active and standby supervisors. When a switchover occurs, the standby supervisor becomes active and uses the synchronized state information to continue forwarding traffic immediately. VRRP (Virtual Router Redundancy Protocol) and HSRP (Hot Standby Router Protocol) are first-hop redundancy protocols that provide gateway redundancy for end devices, not supervisor engine redundancy within a chassis. NSF/NSR (Nonstop Forwarding with Nonstop Routing) works in conjunction with SSO to ensure routing protocols re-converge quickly after a supervisor switchover, but SSO is the primary feature that enables the switchover itself without disrupting forwarding.