18.13 HANA, Cloud & Non-SAP Integration Playbook
Discovery
- Source owner identified.
- Source technology identified.
- Data volume measured.
- Freshness SLA defined.
- Consumer workload understood.
- Security classification documented.
- Source constraints documented.
Pattern selection
- Federation evaluated.
- Replication evaluated.
- Data flow / ETL evaluated.
- Open SQL push evaluated.
- Existing enterprise integration platform evaluated.
- Duplicate persistence challenged.
Engineering
- Initial load defined.
- Delta strategy defined.
- Delete handling defined.
- Idempotency defined.
- Schema evolution defined.
- Reconciliation defined.
- Recovery defined.
Performance
- Remote-query workload measured.
- Source protection tested.
- Statistics evaluated where relevant.
- Filtering applied.
- Unnecessary columns removed.
- Persistence used where justified.
- Concurrency tested.
Security
- Technical identities separated.
- Least privilege implemented.
- Credentials or certificates lifecycle managed.
- Network path documented.
- Audit requirements defined.
- Data access controls validated.
Enterprise architecture
- System of record identified.
- Integration ownership identified.
- Semantic ownership identified.
- Data product boundary identified.
- Failure domains identified.
- Cost model identified.
- Retirement strategy identified.
Architect interview framework
When asked:
"How would you integrate a non-SAP cloud platform with Datasphere?"
Answer:
- Understand source ownership and workload.
- Define freshness, volume and latency requirements.
- Evaluate federation, replication, ETL and push.
- Validate the connection's actual supported features.
- Design security and network boundaries.
- Design initial and delta processing.
- Define reconciliation and recovery.
- Measure source and target performance.
- Establish semantic ownership.
- Expose governed data products.
- Define cost and operational controls.
- Document the retirement or change strategy.
Final architect test
You should be able to defend:
- why data is federated or replicated,
- why persistence exists,
- where transformation runs,
- how the source is protected,
- how failures are recovered,
- how freshness is measured,
- how security crosses boundaries,
- how cloud platforms coexist,
- how duplicate data is controlled.
The objective is not maximum connectivity.
The objective is a reliable, governed and economically sustainable enterprise integration architecture.