What are the DP-600 domain weights?

Maintain a data analytics solution is 25-30%, Prepare data is 45-50%, and Implement and manage semantic models is 25-30%.

Which query languages should a DP-600 candidate know?

The current audience profile calls for the ability to query and analyze data with SQL, KQL, and DAX.

Does DP-600 cover security and deployment?

Yes. Objectives include workspace, item, row, column, object, and file access; sensitivity and endorsement; Git, PBIP, deployment pipelines, impact analysis, XMLA, and reusable assets.

Are PrepKloud DP-600 questions copied from the exam?

No. PrepKloud creates original scenarios from public objectives and official Microsoft documentation and does not use dumps or recalled live questions.

DP-600 Microsoft Fabric Analytics Engineer Roadmap

Q: Is DP-600 active in August 2026?

Yes. DP-600 is active as of August 19, 2026. The current English skills measured took effect July 21, 2026; verify the official Microsoft study guide before booking.

Prepare for Implementing Analytics Solutions Using Microsoft Fabric by connecting governance, data preparation, SQL/KQL/DAX analysis, enterprise semantic modeling, performance, and release engineering. The five phases follow the official skills measured from July 21, 2026.

Exam: DP-600Fabric Analytics Engineer AssociateFive phasesSQL · KQL · DAX

Current scope: DP-600 is active as of August 19, 2026. Microsoft updates role-based exams; verify the official study guide and the date of your localized exam before scheduling. PrepKloud content is independently written from public objectives and documentation—never from dumps or recalled live questions.

Official domain balance

Prepare data carries the largest range, but Fabric solutions cross boundaries. A Direct Lake performance problem might originate in tiny Delta files; an RLS design can fail if the user also has a direct SQL path; and a correct semantic-model change can still break downstream reports if it skips impact analysis.

25-30% · Maintain a data analytics solutionWorkspace/item security, row/column/object/file controls, labels, endorsement, Git, PBIP, deployment pipelines, impact analysis, XMLA, and reusable assets.

45-50% · Prepare dataConnections, OneLake catalog, Real-Time hub, stores, OneLake integration, transformation, star schemas, SQL, KQL, DAX, and visual queries.

25-30% · Implement and manage semantic modelsStorage modes, relationships, enterprise DAX, calculation groups, dynamic formats, field parameters, composite models, Direct Lake, incremental refresh, and optimization.

Fabric foundation, discovery, and store decisions

Weeks 1-2

Begin with workload requirements and the data path. Learn where OneLake fits, how users discover batch and streaming data, and why lakehouse, warehouse, and eventhouse serve different analytical shapes.

Explain OneLake as the tenant's unified logical data lake and distinguish item storage from workspace organization.

Use OneLake catalog for discoverable governed data and Real-Time hub for discoverable streams.

Choose lakehouse for Spark/mixed-data lake engineering, Warehouse for SQL-first structured analytics, and Eventhouse for high-volume events and KQL.

Select among copy, pipeline, Dataflow Gen2, notebook, mirroring, upload, Eventstream, and OneLake shortcut access.

Explain how Eventstream, Eventhouse, KQL database, KQL queryset, real-time dashboard, Activator, and OneLake availability connect.

Create a connection without embedding credentials, document identity behavior, and test source authorization.

Transform, model, and query prepared data

Weeks 3-5

This phase receives the most time because Prepare data is 45-50%. Build reproducible medallion layers, publish dimensional structures, and use the query language native to each engine.

Design bronze raw evidence, silver validated/conformed Delta tables, and gold business-ready facts, dimensions, and aggregates.

Resolve duplicates, missing values, nulls, invalid types, late data, and referential failures with deterministic rules and quarantine evidence.

Implement joins, merges, filters, enrichment, denormalization, aggregation, views, functions, and stored procedures where supported.

Build a lakehouse or warehouse star schema with declared fact grain, conformed dimensions, surrogate keys, and reconciliation.

Use SQL for relational analysis, KQL where/summarize/bin for event analysis, DAX for filter-context calculations, and Visual Query Editor for low-code shaping.

Optimize Delta table file sizes and evaluate V-Order from measured read-versus-write requirements.

Build enterprise semantic models

Weeks 6-7

Model for correctness first, then scale. Understand storage mode and relationship behavior before writing advanced DAX or adding report features.

Choose Import, DirectQuery, Direct Lake, Dual, or composite patterns from freshness, scale, source, feature, security, and capacity constraints.

Distinguish Direct Lake on OneLake from Direct Lake on SQL analytics endpoint and explain fallback, refresh/framing, and default behavior.

Create semantic-model star schemas, role-playing dimensions, bridge tables, and intentional filter directions without ambiguity.

Write DAX with variables, iterators, table filters, windowing, information functions, and explicit context reasoning.

Use calculation groups, dynamic format strings, and field parameters to reduce duplication while preserving numeric behavior.

Configure incremental refresh and large semantic model storage format for the correct problem rather than as interchangeable scale switches.

Secure, govern, and operate the analytics product

Weeks 8-9

Security depends on the access path. Trace workspace, item, OneLake, SQL, and semantic-model permissions and test with the identities users actually have.

Apply least-privilege workspace roles and item sharing; distinguish Read, Build, Reshare, Write, and compute/data permissions.

Implement and test row-, column-, object-, table/folder-, and file-level controls at the required enforcement layer.

Explain why model RLS/OLS does not secure direct lakehouse, Warehouse, Spark, or KQL paths and why Contributor-level roles alter assumptions.

Apply sensitivity labels and endorsement, define owners, review lineage, and record residual access and export risks.

Use Monitoring hub, query history/insights, refresh history, Direct Lake processing evidence, and Capacity Metrics for incident diagnosis.

Optimize KQL reduction, SQL queries, DAX, visual count, model cardinality, Delta layout, workload schedules, and capacity from evidence.

Lifecycle, projects, and exam readiness

Weeks 10-11

Finish by making change safe. Version definitions, test dependencies and security in nonproduction, deploy through stages, and explain decisions under fresh scenarios.

Use PBIP for source-control-friendly Power BI development and distinguish .pbit, .pbids, and shared semantic-model reuse.

Configure supported Fabric Git integration, branches, reviews, synchronization, and secret-safe repository content.

Promote supported items through development, test, and production deployment-pipeline stages with environment configuration and approvals.

Perform lineage and impact analysis, deploy/manage models through XMLA where appropriate, and test rollback.

Complete both DP-600 projects with quality, security, performance, cost, capacity, deployment, validation, and cleanup evidence.

Take fresh mixed-domain practice, explain every distractor, revisit weak objectives, and recheck the current official guide.

PrepKloud learning surfaces

Original practice

Use 25 scenarios balanced to the current domain ranges and explain the governing constraint before checking the answer.

Configure DP-600 practice →

Retrieval flashcards

Recall service boundaries, query patterns, security layers, Direct Lake behavior, and lifecycle distinctions without answer choices.

Open DP-600 flashcards →

Portfolio projects

Build a governed retail lakehouse model and a Real-Time/warehouse executive release lifecycle.

Explore DP-600 projects →

Study guide

Read the long-form domain strategy, architecture decisions, performance loop, project approach, and readiness framework.

Read the DP-600 guide →

Career research

Compare repeated Fabric analytics requirements in current job descriptions with evidence from your projects.

Explore analytics roles →

Official Microsoft references

DP-600 study guide

The authoritative current domain ranges, objectives, audience profile, and change log.

Open the study guide →

Fabric documentation

The official entry point for OneLake, engineering, Warehouse, Real-Time Intelligence, Power BI, governance, and administration.

Open Fabric docs →

Direct Lake overview

Review storage variants, use cases, model behavior, fallback, refresh, security, and limitations.

Study Direct Lake →

Fabric CI/CD

Ground Git integration and deployment pipelines in current item support and release behavior.

Study Fabric CI/CD →

Frequently asked questions

Is DP-600 active in August 2026?

Yes. DP-600 is active as of August 19, 2026. The English skills measured changed July 21, 2026. Localized exams can update later, so verify the guide and exam details for the language and date you choose.

How should study time be allocated?

Give approximately half to data preparation, then split the remainder between solution maintenance and semantic models. Include mixed exercises because security, storage, query performance, and deployment decisions interact.

Must I know SQL, KQL, and DAX?

Yes. The current role profile explicitly expects candidates to query and analyze with all three. Practice selecting, filtering, aggregating, joining, and diagnosing performance in each language's appropriate engine.

Is Power BI knowledge alone enough?

No. Semantic modeling and DAX are important, but the current exam also covers Fabric data stores, OneLake and Real-Time discovery/integration, SQL/KQL preparation, governance, security, Git, PBIP, deployment pipelines, impact analysis, and XMLA.

Does PrepKloud reproduce live exam questions?

No. The practice set is original educational content grounded in the public study guide and official Microsoft documentation. Avoid dumps and recalled questions; they undermine both exam integrity and transferable engineering skill.

Turn every objective into evidence

Read the official behavior, retrieve it from memory, make a design decision, build a synthetic test, inject one failure, and record the security, performance, and capacity result.

Start practice Build projects Read the guide Explore jobs