The secondaries replicate the primary replica set’s oplog and apply the operations to their datasets such that the secondaries’ datasets reflect the primary’s dataset.
