With cross-region DR (disaster recovery) you can create a replica of a vCore-based Azure Cosmos DB for MongoDB cluster in another region. This cluster replica is continuously updated with the data written in the primary region. In a rare case of an outage in the primary region and primary cluster unavailability, you can promote this replica to become the new read-write cluster in another region.
In addition, you can use cluster replica in another region for reading. This helps offload read operations from the primary cluster as well as deliver reads with lower latency to applications that are hosted in or closer to the other region.
When the cluster replica in region B is promoted and becomes the new read-write cluster, it mirrors the original configuration by maintaining a read replica in the former primary region A. Connection strings pointing to the read-write cluster and read-only cluster replica are preserved after replica promotion.
