Enterprise Applications and Data Management
Enterprise Applications and Data Management tests how business applications and data are integrated, protected, and recovered, with emphasis on service levels and data quality rather than programming.
Core rules
- ERP integrates finance, supply chain, and HR; CRM manages customers; HRIS manages employee records. Integration commonly uses APIs, message queues, or batch files.
- Database transactions follow ACID: atomicity, consistency, isolation, and durability.
- Normalization reduces duplication; denormalization can improve read performance at the cost of consistency.
- RPO is acceptable data loss; RTO is acceptable downtime. Backup frequency supports RPO; recovery procedures support RTO.
- Master data needs an owner, definitions, and quality rules. ETL extracts, transforms, and loads; it is not automatically live integration.
- SaaS shifts infrastructure management, not data ownership or access governance.
Worked example
Finance requires no more than 15 minutes of lost transactions and recovery within four hours. RPO is 15 minutes, so transaction log backups or replication must run at least every 15 minutes. RTO is four hours, so the recovery runbook, standby capacity, and restore test must meet four hours. Daily full backups alone cannot meet a 15-minute RPO.
Traps
Do not reverse RPO and RTO. Do not treat replication as backup; deleted or corrupted data replicates too. Do not assume API integration is real-time or reliable without retries and idempotency. Do not load unvalidated data into production master records.