Maintain Analytics Solution
25-30%of exam
Prepare Data
45-50%of exam
Manage Semantic Models
25-30%of exam
Quick Facts
- Exam
- DP-600
- Credential
- Fabric Analytics Engineer
- Time
- 100 min
- Pass
- 700/1000
- Level
- Intermediate
- Product
- Microsoft Fabric
- Focus
- Semantic + data
- Blueprint
- Apr 20 2026
Access Layers
Workspace, item, file, row, column, object
RLS vs CLS
RLS
- Filters rows
- User scoped
CLS
- Hides columns
- Schema scoped
Rows need RLS
Security Picker
- Need workspace authors→Contributor(Least privilege)
- Need membership control→Member(Broad role)
- Need settings control→Admin
- Share one item→Item access
- Filter rows→RLS
- Hide columns→CLS
- Hide tables→OLS
- Protect lake folders→File security
Security + Governance
- Admin
- Full workspace control
- Member
- Manage users/content
- Contributor
- Create/edit content
- Viewer
- Read workspace content
- Item access
- Single-item permission
- RLS
- Row filtering
- CLS
- Column filtering
- OLS
- Object hiding
- File security
- Path-level access
- Sensitivity label
- Information classification
- Endorsement
- Trusted content signal
- Lineage
- Dependency map
Item vs Workspace
Item access
- Single asset
- Least exposure
Workspace role
- All workspace items
- Broader exposure
Share narrowly first
Lifecycle
- Git integration
- Workspace versioning
- Branch
- Isolated changes
- Commit
- Saved snapshot
- Sync
- Align workspace/Git
- PBIP
- Power BI project
- Pipeline
- Dev-test-prod promotion
- Deployment rule
- Stage binding override
- Impact analysis
- Downstream dependencies
- XMLA endpoint
- Model management endpoint
- Workspace settings
- Environment configuration
Reusable Assets
- PBIT
- Report template
- PBIDS
- Data source shortcut
- Shared model
- Reusable semantic layer
- Certified model
- Approved enterprise source
- Promoted item
- Recommended content
- Template app
- Packaged solution
- Parameter
- Environment input
- Connection
- Reusable credential binding
Store Trio
Lake files, warehouse SQL, eventhouse KQL
Lakehouse vs Warehouse
Lakehouse
- Delta files
- Spark friendly
Warehouse
- Relational SQL
- T-SQL friendly
Pick by workload
Store Picker
- Open Delta files→Lakehouse
- Relational SQL analytics→Warehouse
- Real-time telemetry→Eventhouse
- Reuse external data→Shortcut
- Low-code shaping→Dataflow Gen2
- Custom Spark code→Notebook
- Business semantic layer→Semantic model
- Discover certified data→OneLake catalog
- Streaming sources→Real-Time hub
Store Selection
- Lakehouse
- Open Delta analytics
- Warehouse
- Relational SQL analytics
- Eventhouse
- Real-time KQL analytics
- KQL database
- Telemetry/time-series store
- OneLake
- Unified data lake
- Shortcut
- Virtual data reference
- Mirroring
- Near-real-time replica
- Semantic model
- BI consumption layer
- Real-Time hub
- Streaming discovery
- OneLake catalog
- Data discovery
Eventhouse vs Warehouse
Eventhouse
- Streaming telemetry
- KQL queries
Warehouse
- Structured tables
- T-SQL queries
Telemetry favors KQL
Data Ingestion
- Connection
- Source access binding
- Copy activity
- Move data
- Dataflow Gen2
- Low-code shaping
- Notebook
- Spark code
- Pipeline
- Activity orchestration
- Eventstream
- Event routing
- Full load
- Reload all rows
- Incremental load
- Load changes only
- CDC
- Change capture
- Gateway
- On-prem connectivity
Dataflow vs Notebook
Dataflow Gen2
- Low-code Power Query
- Visual transforms
Notebook
- Code transforms
- Spark control
Code needs notebooks
Transform Data
- View
- Reusable query
- Function
- Reusable logic
- Stored procedure
- SQL routine
- New column
- Enrichment field
- New table
- Enrichment output
- Denormalize
- Flatten structure
- Aggregate
- Summarize rows
- Merge
- Combine columns
- Join
- Relate tables
- Filter
- Keep rows
- Type conversion
- Fix data types
- Star schema
- Facts plus dimensions
Query Tools
- Visual Query
- Graphical select/filter
- SQL
- Relational querying
- KQL
- Telemetry querying
- DAX
- Model calculations
- SELECT
- Choose columns
- WHERE
- Filter rows
- GROUP BY
- Aggregate groups
- summarize
- KQL aggregate
- CALCULATE
- Modify filter context
- EVALUATE
- Return DAX table
Data Quality
- Duplicate rows
- Deduplicate by key
- Missing data
- Impute or reject
- Null values
- Handle blanks
- Bad types
- Convert columns
- Schema drift
- Source changed
- Outlier
- Suspicious value
- Late data
- Delayed arrival
- Surrogate key
- Warehouse identifier
- Business key
- Source identifier
- Validation rule
- Quality check
STAR Model
Facts measure, dimensions describe
Import vs Direct Lake
Import
- Copied cache
- Scheduled refresh
Direct Lake
- Reads OneLake
- Fast freshness
OneLake enables Direct Lake
Model Picker
- Need import speed→Import
- Need live source→DirectQuery
- Need OneLake speed→Direct Lake
- Mix source modes→Composite model
- Huge model→Large format
- Role-based rows→RLS
- Reusable measure logic→Calc group
- User dimension switch→Field parameter
Model Design
- Import
- Cached model data
- DirectQuery
- Live source queries
- Direct Lake
- OneLake table reads
- Composite model
- Mixed storage modes
- Fact table
- Measures/events
- Dimension table
- Descriptive context
- Bridge table
- Many-to-many resolver
- Relationship
- Table filter path
- Cardinality
- Row matching shape
- Filter direction
- Propagation path
DAX VAR
Store once, reuse often
Direct Lake vs DirectQuery
Direct Lake
- Parquet reads
- Fallback possible
DirectQuery
- Source queries
- Latency sensitive
Fallback behaves DirectQuery
DAX Patterns
- Measure
- Dynamic calculation
- Calculated column
- Stored row value
- VAR
- Reusable expression
- Iterator
- Row-by-row evaluation
- FILTER
- Table filtering
- WINDOW
- Window calculation
- INFO
- Metadata function
- Context transition
- Row becomes filter
- Time intelligence
- Date-aware measures
- DIVIDE
- Safe division
Calc Group vs Field Parameter
Calc group
- Measure variants
- Format logic
Field parameter
- Field switching
- User selection
Measures need calc groups
Semantic Features
- Calc group
- Reusable measure variants
- Calc item
- Group expression
- Dynamic format
- Measure-specific format
- Field parameter
- User field switch
- Large format
- Large model storage
- Aggregation table
- Pre-summarized data
- Perspective
- Field subset
- Hierarchy
- Drill path
- Display folder
- Field organization
- XMLA write
- External model changes
Optimization + Refresh
- Fallback
- DirectQuery downgrade
- Framing
- Direct Lake eligibility
- Refresh
- Model data update
- Incremental refresh
- Partitioned updates
- Query reduction
- Fewer visual queries
- Performance Analyzer
- Visual timing trace
- DAX Studio
- Query performance tool
- VertiPaq
- Import storage engine
- Cardinality
- Compression pressure
- Star schema
- Fast filter design
Common Traps
Viewer is not builder
Viewer reads ≠ Contributor builds
Sensitivity is not security
Label classifies ≠ Access controls
Endorsement is not deployment
Endorse trusts ≠ Pipeline promotes
Direct Lake is not import
Reads OneLake ≠ No copied cache
Warehouse is not eventhouse
T-SQL relational ≠ KQL telemetry
RLS is not CLS
Rows filtered ≠ Columns hidden
PBIT is not PBIDS
Template report ≠ Connection shortcut
Last Minute
- 1.Map three domain weights
- 2.Pick least privilege role
- 3.Separate item from workspace
- 4.Use RLS for rows
- 5.Use CLS for columns
- 6.Choose correct Fabric store
- 7.Query with SQL KQL DAX
- 8.Prepare star schemas
- 9.Prefer variables in DAX
- 10.Know Direct Lake fallback
- 11.Review deployment pipeline rules
- 12.Check lineage before changes
Explore More Microsoft Azure Certifications
Continue into nearby exams from the same family. Each card keeps practice questions, study guides, flashcards, videos, and articles in one place.
More From This Family
Videos and articles for deeper review.
