BCP (business continuity plan): IT does not stop. An active backup site takes over immediately — continuous replication, automatic or near-automatic failover. The goal is “zero perceived interruption”, at the price of duplicated infrastructure and a permanent inter-site link.
DRP (disaster recovery plan): IT restarts after the disaster, from backups or asynchronous replication, within a target delay (RTO) of hours to days. Cheaper, but you accept the interruption.
Mature architectures often combine both: BCP on a nearby data center (synchronous replication, same metro) for common failures, DRP on a distant site (separate risk zone) for major disasters. Both require distinct data centers and regular failover testing — an untested plan does not exist.