How should Azure availability and disaster recovery be designed for production in Microsoft Azure?

Define workload RTO/RPO first, map dependencies, use zone redundancy where appropriate, automate backups, test failover, document runbooks, and validate DNS and data recovery procedures.