Microsoft Certified: Fabric Analytics Engineer Associate — DP-600
Complete Study Course
Exam: DP-600, Implementing Analytics Solutions Using Microsoft Fabric
Audience: Analytics engineers who turn data into reusable analytics assets in Microsoft Fabric: lakehouses, warehouses, notebooks, dataflows, pipelines, semantic models and reports. You should be comfortable with SQL, DAX and PySpark, data modeling, and Git-based source control.
Domain Blueprint
| Domain | Weight | Lessons | Focus |
|---|---|---|---|
| 1. Maintain a data analytics solution | 25–30% | 7 | Capacity, settings and gateways; access, sharing and granular security; labels and endorsement; Git and .pbip; deployment pipelines; XMLA and reusable assets |
| 2. Prepare data | 45–50% | 9 | Discovery and connections; ingestion and shortcuts; lakehouse, warehouse or eventhouse; views and procedures; star schema, SCD and bridge tables; cleansing; SQL and KQL querying; load performance |
| 3. Implement and manage semantic models | 25–30% | 8 | Storage modes and Direct Lake; relationships; DAX; calculation groups and field parameters; large and composite models; RLS and OLS; optimization tools; incremental refresh |
On Skillmarked: 24 free lessons and 24 practice questions, with a practice scenario after every lesson.
Note on the outline: Microsoft's DP-600 pages don't currently agree. The certification page lists the three domains above. The study guide page still shows an older four-domain outline (Plan, implement and manage a solution; Prepare and serve data; Implement and manage semantic models; Explore and analyze data). This course follows the three current domains and also covers the older outline's extra topics: capacity SKUs, gateways, report themes, the visual query editor and exploratory analytics. Check Microsoft's pages before you book.
Recommended Study Plan (30–40 hours with Power BI experience)
Week 1 – Prepare data (Domain 2, the biggest)
- Build a lakehouse and a warehouse, and load them with a pipeline, a Dataflow Gen2 dataflow and a notebook
- Write a Type 2 dimension load in T-SQL and in PySpark, and model a bridge table
- Practice WHERE vs HAVING, window functions and short KQL queries
Week 2 – Semantic models (Domain 3)
- Build a Direct Lake model and an Import model on the same gold tables, and compare framing with refresh
- Write measures with variables, CALCULATE and ALLSELECTED, OFFSET and WINDOW, then a calculation group and a field parameter
- Implement dynamic RLS with USERPRINCIPALNAME and test it with View as
Week 3 – Run it like production (Domain 1)
- Save a report as .pbip, commit it to Git, and deploy through a deployment pipeline with rules
- Connect DAX Studio and Tabular Editor through the XMLA endpoint
- Take timed mock exams and go back over your weakest domain
Critical Concepts (High-Yield)
Direct Lake
- Framing (a seconds-long metadata refresh) instead of a data copy. Only Direct Lake on SQL falls back to DirectQuery (views, SQL RLS, guardrails), controlled by DirectLakeBehavior
DAX filter control
ALLignores filters,ALLSELECTEDkeeps outer selections,KEEPFILTERSadds without overriding, andUSERELATIONSHIPturns on an inactive relationship
Security layers
- Workspace roles and item permissions (Build for model reuse), warehouse GRANT, DENY and RLS, OneLake security roles, and model RLS and OLS. RLS doesn't restrict Admins, Members or Contributors
Lifecycle
- .pbip = TMDL plus PBIR text files for Git. Deployment pipelines copy definitions, not data or refresh schedules. XMLA is read-only until a capacity admin sets it to Read Write
The exam is open book: Microsoft Learn is available during role-based exams, but the clock keeps running.