A company runs critical workloads in an on-premises data center.

AWS Elastic Disaster Recovery (AWS DRS)is purpose-built for continuously replicatingphysical and virtual serversinto AWS with low RTO/RPO. It uses lightweight replication agents to stream block-level changes to a low-coststaging areain AWS, which helps optimize storage costs (only the staging resources run continuously) and reduces bandwidth usage through efficient replication mechanisms. In a disaster or test, AWS DRS can automatically launch recovery instances in AWS based on a defined blueprint (instance types, networking, security groups), enabling rapid failover workflows that commonly meetsub-hour RTOobjectives.

Option A is not the intended service model: AWS Backup protects AWS-native resources and does not “directly replicate” arbitrary on-prem servers as a continuous replication DR system. Option B (Storage Gateway Volume Gateway) can support backups of certain storage use cases via snapshots, but it is not a general continuous replication solution for diverse physical/virtual servers and may not meet the RTO requirement as directly as AWS DRS. Option D (Direct Connect + custom automation) can help with connectivity, but it does not provide continuous server replication by itself and would require significant custom engineering and ongoing operational effort.

Therefore, enabling AWS Elastic Disaster Recovery and configuring replication agents is the best automated, cost-optimized solution.