Skip to main content

Module 17: Open ODS Views (Advanced) – Virtual & Federated Modeling (BW/4HANA 2.0)

Open ODS Views are a cornerstone of virtualization and data federation in BW/4HANA.
They enable fast integration of external data without physical persistence, fully leveraging HANA pushdown.

This module covers:

  • Virtual data models
  • External data integration
  • Open ODS vs aDSO
  • SQL-based sources
  • Data federation concepts

1. What is an Open ODS View? (Recap + Context)​

An Open ODS View is a virtual BW object that:

  • Exposes external data sources to BW
  • Does not persist data in BW
  • Can be consumed by BW Queries
  • Supports HANA pushdown
Key Principle

Open ODS Views bring data to BW semantics without moving the data.

2. Virtual Data Models (Core Concept)​

2.1 What is a Virtual Data Model?​

A virtual data model:

  • Represents data logically
  • Executes queries at runtime
  • Avoids physical storage
External Source → Open ODS View → CompositeProvider → BW Query

Advantages of Virtual Models​

BenefitExplanation
SpeedNo data load
FlexibilitySchema changes easier
StorageNo BW footprint
Real-timeAlways current
tip

Virtual models are ideal for exploration, prototyping, and real-time use cases.

3. External Data Integration​

3.1 Supported External Sources​

Open ODS Views can integrate:

  • SAP HANA tables & views
  • CDS views
  • External databases
  • Flat files
  • SLT-replicated tables

Typical Use Cases​

  • Side-by-side analytics
  • Legacy system access
  • Cloud data federation
  • Rapid integration projects
warning

Open ODS Views are not meant for heavy transformations.

4. Open ODS View vs aDSO (Very Important)​

Comparison Table​

AspectOpen ODS ViewaDSO
PersistenceNoYes
Data LoadNoYes
PerformanceRuntime-basedLoad-time optimized
HistoryNoYes
Delta HandlingLimitedFull
Use CaseFederationCore DW

Architectural Rule​

Golden Rule

Use Open ODS Views for access
Use aDSOs for storage

5. SQL-Based Sources​

5.1 SQL Views & Tables​

Open ODS Views can be built on:

  • HANA SQL views
  • Calculation views
  • External schemas

Mapping Semantics​

  • Map fields to InfoObjects
  • Define semantics (units, currencies)
  • Enable analytical consumption
info

Semantics are mandatory for BW Query consumption.

Performance Considerations​

  • Push filters down
  • Avoid SELECT *
  • Keep joins simple

6. Data Federation Concepts​

What is Data Federation?​

Data federation means:

  • Querying distributed data sources
  • Without consolidating data physically

Federation in BW/4HANA​

  • Open ODS Views
  • CompositeProviders
  • HANA smart data access (SDA)
Source A ┐
         ├→ Open ODS → CompositeProvider → Query
Source B ┘

When Federation Makes Sense​

  • Real-time requirements
  • Low data volumes
  • Read-only analytics
  • Decentralized ownership
warning

Federation is not a replacement for enterprise data warehousing.

7. Performance Implications (Critical)​

Strengths​

  • No load latency
  • Reduced storage
  • Real-time access

Limitations​

  • Source system dependency
  • Network latency
  • Limited transformation capabilities
Best Practice

Federate selectively, persist strategically.

8. BW/4HANA 2.0 vs 1.0 (Open ODS)​

AspectBW/4HANA 1.0BW/4HANA 2.0
AdoptionLimitedStrong
IntegrationHANA-centricHybrid
PushdownPartialStrong
FederationBasicMature

9. Common Design Mistakes​

Avoid These

Using Open ODS as a replacement for aDSO
Federating very large datasets
Complex joins across systems
Ignoring source system SLAs

10. Interview-Grade Questions​

Q1. When should Open ODS Views be used?​

Answer: When real-time access or external data integration is required and persistence is not necessary.

Q2. Why not use Open ODS Views everywhere?​

Answer: Because they lack historical storage, delta handling, and can introduce performance and availability dependencies.

11. Summary​

  • Open ODS Views enable virtualization
  • Ideal for external and real-time data
  • Complement aDSOs, not replace them
  • Key tool for data federation
  • Must be used selectively

12. What's Next?​

➡️ Module 18: Integration Scenarios & Source Systems

Learning Tip

Virtualization gives speed — persistence gives control.