Your research is right , HA is done using cluster and DR is done using Backup and restore. Reprocessing is not required if the backup is taken after the cube has been processed. You could also perform SAN replication instead of implementing HA but this option doesnt allow failover and needs proper testing.