Provide Solution/Pattern Using Just Databricks for both Streaming & Batch

[adaptiv-rsarangan]

Re. modern analytics architecture using Databricks: https://learn.microsoft.com/en-us/azure/architecture/solution-ideas/articles/azure-databricks-modern-analytics-architecture

The existing solution idea utilises both Databricks and Synapse Analytics for streaming and batch data loads respectively. We have come across customers who would prefer a single platform and given Databricks' capability to handle both streaming and batch processes, it would help to have a reference architecture that purely makes use of Databricks' data warehousing layer.

In the case of link above, Step 8 (Synapse Analytics) is entirely superfluous and the same business use-cases can be achieved without it.

Proposed Solution: Split this into two (or three) solution patterns -

1. One using just Databricks.
2. One as a hybrid of Databricks + Synapse Analytics (current solution).
3. As an alternative, there could also be a third solution idea with just Synapse Analytics.

---

Document Details

⚠ Do not edit this section. It is required for learn.microsoft.com ➟ GitHub issue linking.

• ID: 7254782d-e773-b542-73e8-f568d3bbee39
• Version Independent ID: 40d81223-f8ea-2eec-616e-5f6ee28c8206
• Content: Modern analytics architecture with Azure Databricks - Azure Architecture Center
• Content Source: docs/solution-ideas/articles/azure-databricks-modern-analytics-architecture.yml
• Service: architecture-center
• Sub-service: solution-idea
• GitHub Login: @martinekuan
• Microsoft Alias: architectures

[ManoharLakkoju-MSFT]

@adaptiv-rsarangan
Thanks for your feedback! We will investigate and update as appropriate.

[adaptiv-rsarangan]

Hey team, any updates here? @ManoharLakkoju-MSFT , @martinekuan

Thanks!