SAP HANA DBA Calculations: Sizing, Backup, Memory & Performance Explained

Introduction:
If you are working as an SAP HANA DBA or planning a new system migration, you’ve probably come across the term “calculation”. But what does it actually mean in HANA administration?

In SAP HANA, “calculation” usually refers to estimations and measurements that DBAs perform for sizing, memory usage, backups, recovery planning, and performance analysis. Understanding these calculations helps ensure system stability, faster recovery, and optimized performance.

Click here to read SAP HANA Data Reclaim Process

1. HANA DBA Calculation Sizing (HANA Memory & CPU)

Purpose: Plan hardware requirements for a new or migrated SAP system.

Total Memory = (Data Volume × Compression Factor) + Working Space + System Overhead

• Raw Data: 2 TB
• Compression (4×): 500 GB
• Working Space: 250 GB
• Overhead: 75 GB

Total Memory Required ≈ 825 GB

2. Backup Size Calculation

Purpose: Estimate backup storage for Backint / File-based methods.

• Full Backup ≈ Compressed Data Size
• Differential Backup ≈ Changed Data Since Last Full
• Incremental Backup ≈ Changed Data Since Last Any Backup

Weekly Example: 1 Full (500 GB) + 6 Differentials (600 GB) + 168 Incrementals (840 GB) = ~1.94 TB

3. Memory Usage Calculation (Live Monitoring)

Formula: Used Memory = Column Store + Row Store + Cache + Other Processes

Example: Column: 450 GB, Row: 20 GB, Cache: 30 GB, Other: 15 GB → Total = 515 GB

4. Recovery Calculations (RTO & PITR)

Formula: Log Space = Log Generation Rate × Retention Period

Example: 10 GB/hr × 72 hrs = 720 GB

5. Performance Calculations

• Query Throughput = Rows ÷ Execution Time
• Compression Ratio = Raw Size ÷ Compressed Size

Example: Raw: 2 TB, Compressed: 500 GB → 4:1

Conclusion

These key calculations help SAP HANA DBAs in planning, monitoring, and optimizing the environment. Regular estimations ensure efficiency in memory, backups, recovery, and performance.

Pro Tip: Always revisit your sizing and backup strategy regularly as data volumes and workloads grow.