Refer to the exhibit. Site 1 and site 2 share routes using BGP. OSPF has been implemented within each site to share IP routes. A network engineer is implementing Graceful Restart on R1. Which action…

Enabling NSF globally on all BGP peers at both sites allows them to continue forwarding traffic using the last known routes from R1 even if R1's BGP process restarts (e.g., during an SSO). This feature ensures that routing and forwarding continue uninterrupted during BGP control-plane outages by leveraging stale routes while new BGP sessions are re-established with R1.

Mutually redistributing BGP into OSPF on each router would create routing loops and administrative complexity, and it's not the correct mechanism for maintaining BGP routes during an SSO event.

Extending OSPF between the two sites would merge the IGP domains, which is a different design choice and does not specifically enable BGP peers to retain routes during an R1 BGP restart.

Matching the OSPF timer in each site to the BGP settings is not a direct action that enables BGP peers to use R1's BGP routes during an SSO; it relates to IGP and BGP convergence but not graceful restart capabilities.