Black Friday50% off your first month on any monthly plan with EARLY50 See plans
All 35 lessons
  1. Start here: cloud, data, Fabric and Power BI basics

    1. 1. Start here: the exam, the plan and what you already know
    2. 2. Cloud in plain words: SaaS, PaaS, IaaS and where Fabric fits
    3. 3. Data basics: files, tables, formats and the lakehouse idea
    4. 4. A tour of Microsoft Fabric: OneLake, workspaces and items
    5. 5. Power BI in plain words: semantic models, reports and DAX
  2. Prepare data: get data in (exam area 2, 45-50%)

    1. 6. Connect to data and find it: connections, gateways, OneLake catalog and Real-Time hub
    2. 7. Choosing a data store: lakehouse, warehouse or eventhouse
    3. 8. Use data without copying: shortcuts, mirroring and OneLake integration
    4. 9. Ingest data: Dataflow Gen2, pipelines, Copy job and notebooks
    5. 10. Power Query in Dataflow Gen2: filter, data types, merge and new columns
    6. 11. Messy data: duplicates, missing values and nulls
  3. Prepare data: transform and query (exam area 2)

    1. 12. Views, functions and stored procedures in T-SQL
    2. 13. Build a star schema in a lakehouse or warehouse
    3. 14. Aggregate, join and enrich: SQL and PySpark side by side
    4. 15. Query with the Visual query editor and SQL
    5. 16. Query with KQL
    6. 17. Query with DAX: DAX query view and EVALUATE
  4. Design and build semantic models (exam area 3, 25-30%)

    1. 18. Choose a storage mode: Import, DirectQuery, Direct Lake and Dual
    2. 19. Relationships: star schemas, bridge tables and many-to-many
    3. 20. DAX foundations: measures, filter context, CALCULATE and variables
    4. 21. DAX functions the exam tests: iterators, table filters, windows and information functions
    5. 22. Calculation groups, dynamic format strings and field parameters
    6. 23. Composite models and large semantic model storage format
  5. Optimize models, then secure them (exam areas 3 and 1)

    1. 24. Direct Lake in depth: fallback, refresh and the two flavors
    2. 25. Incremental refresh for semantic models
    3. 26. Faster reports and DAX: finding and fixing slow queries
    4. 27. Who can see what: workspace roles and item permissions
    5. 28. Fine-grained security: rows, columns, objects and files
    6. 29. Governance: sensitivity labels and endorsement
  6. Lifecycle, then exam practice (exam area 1)

    1. 30. Version control: Git integration and Power BI projects (.pbip)
    2. 31. Deployment pipelines and impact analysis
    3. 32. XMLA endpoint and reusable assets: .pbit, .pbids and shared semantic models
    4. 33. Mock exam 1
    5. 34. Final review and exam day: every decision on one page
    6. 35. Mock exam 2
See the leaderboard

24 min

Final review and exam day: every decision on one page

Every decision rule the exam tests, condensed with the lesson that teaches it, plus a timed method for exam day.

By the end of this lesson you can

  • Recall the decision rule for every choose-between question in the course
  • Name the facts the exam expects you to know cold, with the lesson behind each
  • Budget 100 minutes across standalone questions, the case study and review
  • Use the constraint word to pick between answers that would all work

From earlier lessons · 1 of 3

You're about to rename a column in Juniper Coffee's sales semantic model. Reports in other teams' workspaces may use it. How do you find them and warn their owners?

Pick one answer.

From earlier lessons · 2 of 3

Tabular Editor connects to Juniper's workspace and reads the model, but saving a new calculation group fails. What should change?

Pick one answer.

From earlier lessons · 3 of 3

You try to commit your report changes from Juniper's dev workspace, but Commit is blocked. The Source control pane lists items under Updates. What do you do first?

Pick one answer.

You're the analytics engineer at Juniper Coffee, and mock exam 1 is marked: 64%. Most misses weren't gaps in knowledge. You picked Import when the data was already in OneLake. You gave a Contributor role when one item was enough. The exam rewards fast, sure choices, and those come from rules you can recall in seconds. This page puts every rule from the course in one place.

[6]

How to use this page in 20 to 25 minutes. Each table condenses one decision from an earlier lesson. The (Ln) tag names the lesson that teaches it in full.

  1. Start with your weakest skill area from mock 1. Prepare data is 45-50% of the exam; the other two areas are 25-30% each.
  2. Cover the right-hand column of each table and say the answer out loud.
  3. Missed a row? Reopen that lesson's flashcards, then its Exam-style practice.

Readiness rule: book the exam when you score 80% or more on both mock exams. The mocks pass at 70%. The real score is scaled, so a margin above the line protects you.

Prepare data: where data lives and how it gets in

Exam depth: know it cold.

Which data store (L7)

The scenario says.Pick.
Streaming events, telemetry, logs, answers within seconds, KQL.Eventhouse.
A T-SQL team that writes, multi-table transactions, stored procedures.Warehouse.
PySpark, notebooks, files, unstructured or mixed data, or not sure.Lakehouse.
T-SQL users who only read.Lakehouse with its SQL analytics endpoint.
An operational app, OLTP.SQL database in Fabric.
NoSQL, documents.Cosmos DB in Fabric.

Ingest or access (L8)

The scenario says.Pick.
No data movement, the data stays where it is.Shortcut.
A live, read-only copy of an outside database, no ETL.Mirroring.
KQL database data needed in OneLake as Delta.OneLake availability.
An Import model's tables needed as Delta.OneLake integration for semantic models.
A schedule, transforms on the way, or a writable table.A copy: pipeline, Copy job or dataflow.

A Tables shortcut to Delta data becomes a table the SQL analytics endpoint and Direct Lake can read. A Files shortcut suits raw files, such as CSV, that Spark reads by path. Access through a shortcut is the most restrictive of the two permissions. OneLake availability can lag up to 3 hours by default.

[81][82][83][84][87][88][89]

Which ingest tool (L9)

The scenario says.Pick.
No code, analysts reshape data on the way in.Dataflow Gen2.
A plain copy, full or incremental, no pipeline to build.Copy job.
A copy as one step in an ordered, scheduled flow.Pipeline with a Copy activity.
Large or messy data, a PySpark team.Notebook.
A SQL team loading Warehouse tables.T-SQL: COPY INTO, INSERT ... SELECT, CTAS.

Dataflow Gen2 destinations: Replace for a full file each time, Append for new rows only. Append plus a full file means duplicates. Fixed schema keeps relationships and measures. A Warehouse destination needs staging on, fixed schema and the same workspace. In a Copy job, a watermark catches inserts and updates; CDC also catches deletes.

Which gateway (L6)

The source is.Pick.
In your own building, behind a firewall.On-premises data gateway.
In Azure, behind a private endpoint or virtual network.Virtual network data gateway.
A public cloud service.No gateway.

A workspace identity signs in with no secrets to manage. It can't live in My workspace or work on gateway connections.

[68][70][71][90][92][94][96]

You are an analytics engineer at Juniper Coffee. Till sales live in an Azure SQL Database. Analysts need a near real-time, read-only copy in OneLake to query with T-SQL. Nobody wants to build or maintain an ETL job. What should you use?

Pick one answer.

[40][83][87]

Prepare data: shape, clean and query

Exam depth: know it cold.

Merge or append, and the join kind (L10)

You need.Use.
More columns from a second table, matched on a key.Merge queries (a join).
More rows from tables with the same column names.Append queries (a union, matched by name).
All left rows, plus matches (the default).Left outer.
Only rows that match on both sides.Inner.
All rows from both tables.Full outer.
Left rows with no match, such as products nobody prices.Left anti.

Key columns must share a data type, or the merge returns wrong results.

Messy data (L11)

Problem.Power Query.T-SQL.PySpark.
Identical copies.Remove duplicates.SELECT DISTINCT.dropDuplicates().
Keep the newest row per key.Not reliable.ROW_NUMBER() ordered DESC, keep 1.row_number() over a window.
Null to a label.Replace values.COALESCE or ISNULL.fillna.
A value that won't convert.Error cell.CAST fails; TRY_CAST gives NULL.cast gives null.

Remove duplicates is case-sensitive. Fill down works only on nulls. Profiling reads the top 1,000 rows unless you switch it. Never drop a sale with a bad key: point it at an Unknown member, such as key -1.

[34][98][103][104][113][121][123][125]

View, function or procedure (L12)

You need.Use.
The same rows for everyone, shared logic or a security layer.View.
Rows filtered by a value you pass in, used in FROM.Inline table-valued function.
One value per row, a reusable calculation.Scalar function.
Steps that change data, a repeatable load.Stored procedure, run with EXEC or a pipeline.

Use CREATE OR ALTER so grants survive. A SQL analytics endpoint takes views, functions and read-only procedures, but no CREATE TABLE, CTAS, INSERT, UPDATE, DELETE or MERGE. Triggers and T-SQL materialized views aren't supported in Fabric Warehouse.

Which editor (L15): no SQL skills, the Visual query editor (read-only SELECT, steps that fold). Anything that changes data, the SQL query editor.

Star schema (L13): flatten snowflaked lookups into one dimension. SCD Type 2 when the past must stay as it was; Type 1 overwrites. Load dimensions first. Fabric IDENTITY is BIGINT, unique but not sequential.

Read the query (L14, L16): WHERE filters rows, HAVING filters groups. On 700, 700, 500: RANK gives 1, 1, 3 and DENSE_RANK gives 1, 1, 2. In KQL, sort and top are descending by default, take returns any rows, and where after summarize acts as HAVING.

[34][123][139][142][143][156][168][175]

You are an analytics engineer at Juniper Coffee. Analysts write many T-SQL queries on the Warehouse. Each query needs sales rows for one region that the analyst names, joined to other tables in FROM. You need one reusable object. What should you create?

Pick one answer.

[139][143]

Semantic models: storage modes and Direct Lake

Exam depth: know it cold.

Storage mode (L18)

The scenario says.Pick.
Data already in OneLake Delta tables, Fabric capacity, no refresh window.Direct Lake.
Data must stay in a non-Fabric source, or be near real-time from it.DirectQuery.
Small or slow-changing data, Power Query needed, or Pro licenses only.Import.
A dimension shared with DirectQuery facts from the same source.Dual.

Import is fastest but only as fresh as the last refresh: 8 scheduled refreshes a day on shared capacity, 48 on Premium or Fabric. Direct Lake needs a Fabric capacity, and its refresh is framing: metadata only, in seconds. A table switched to Import can't go back to DirectQuery.

Composite models and large format (L23)

The scenario says.Do this.
Extend a shared model you can't edit, with no copy.DirectQuery to the semantic model, then add local tables.
Live DirectQuery facts, fast slicers, same source.Set the dimensions to Dual.
Rows vanish or RELATED fails across sources.It's a limited relationship.
A Direct Lake model plus an Excel table.Direct Lake on OneLake plus Import, in web modeling.
A model past 10 GB, big incremental refresh, or XMLA writes.Large semantic model storage format.

The small format stops at 1 GB. Large format lifts it to the SKU's memory per model, 25 GB on F64. Turn it on before the first big refresh.

[45][198][201][203][204][240][242]

Direct Lake on OneLake or on SQL analytics endpoint (L24)

Feature.On OneLake.On SQL analytics endpoint.
Sources.One or more Fabric items.One item with a SQL analytics endpoint.
Falls back to DirectQuery.Never.Yes, set by DirectLakeBehavior.
SQL views.Not supported.Allowed, but always fall back.
SQL endpoint RLS.Not applied; queries succeed.Queries fall back, or fail if fallback is off.
Data security.OneLake security roles, or Read and ReadAll.SQL permissions: Read plus SELECT.
Guardrail exceeded.Refresh fails; model can't be queried.Falls back; refresh succeeds with a warning.
Mix with Import tables.Yes, in web modeling.No.
Power BI Desktop creates it.Yes, always.No, only the SQL endpoint page.
Rebind after a pipeline deploy.M parameters with parameter rules.Data source rule.

DirectLakeBehavior (Direct Lake on SQL only): Automatic, the default, falls back silently. DirectLakeOnly makes queries fail instead. DirectQueryOnly always uses DirectQuery.

What makes Direct Lake on SQL fall back: SQL RLS, masking or OLS on a table, a view that isn't materialized, a guardrail exceeded, or tables not framed.

Refresh: automatic updates are on by default. Turn them off when users mustn't see half-loaded data, and refresh as the pipeline's last step. On F64 a table can hold 5,000 Parquet files and 1.5 billion rows. Compact small files with OPTIMIZE.

[45][198][199][306]

You are an analytics engineer at Juniper Coffee. You're testing a Direct Lake on SQL analytics endpoint model before release. Any query that can't run in Direct Lake mode must fail with an error, so you can find and fix the cause. What should you set?

Pick one answer.

[198]

Semantic models: relationships, DAX and refresh

Exam depth: know it cold.

Relationships and bridge tables (L19)

You need.Use.
The default star.One-to-many from dimension to fact, single direction.
Two dimensions that relate many-to-many.A bridge table: one-to-many from each dimension, one relationship set to Both.
A fact at a higher grain, such as budget by region.Many-to-many cardinality, single direction, a measure that blanks lower levels.
Two fact tables.Shared dimensions, never a direct relationship.
A second date role, one at a time.An inactive relationship plus USERELATIONSHIP in a measure.
Several roles at once, or RLS on each path.One copy of the dimension per role.
A slicer that lists only items with data.A visual-level filter: the measure is not blank.
One measure needs filters both ways.CROSSFILTER(..., BOTH) inside that measure.

A one-to-one relationship allows only Both. A duplicate key on the one side fails the refresh. Unmatched fact keys show as a (Blank) row. Many-to-many and cross-source relationships are limited: inner-join logic and no RELATED. RLS travels only on active relationships, and crosses a Both relationship only with Apply security filter in both directions.

[60][205][206][207][209]

DAX function picks (L20, L21)

You need.Write.
Per-row math, such as Qty times UnitPrice.SUMX, an iterator.
A ratio whose denominator may be 0 or blank.DIVIDE.
Percent of total: clear a filter.REMOVEFILTERS in CALCULATE.
Share of a parent: keep only one column's filter.ALLEXCEPT.
Add a filter without replacing the user's.KEEPFILTERS.
Rank against every store, or only the selection.RANKX over ALL, or over ALLSELECTED.
A filter that tests a measure or two tables.FILTER; otherwise a Boolean filter.
Previous month, first month, running total.OFFSET, INDEX, WINDOW.
Behave differently at a level, skip totals.ISINSCOPE.
The one selected value, else a default.SELECTEDVALUE.
Year to date, last year, shift by n periods.TOTALYTD, SAMEPERIODLASTYEAR, DATEADD.

Must react to slicers: a measure. Must slice or group: a calculated column. A VAR is evaluated once, where it's declared. TOPN returns extra rows on ties, unsorted.

Calculation group, field parameter or format string (L22)

You need.Use.
The same logic, such as PY or YoY %, on many measures.Calculation group with SELECTEDMEASURE().
Readers choose which measure or column a visual shows.Field parameter.
Only the display format changes.Dynamic format string.

A calculation group needs Discourage implicit measures on. The highest precedence group wraps outermost, and its format string wins.

[197][214][224][227][229][232][237][239]

Incremental refresh settings (L25)

Setting.What it decides.
RangeStart and RangeEnd.Reserved Date/Time parameters. Filter >= RangeStart and < RangeEnd, and the filter must fold.
Archive data starting ... before refresh date.How much history the model keeps, such as 3 years.
Incrementally refresh data starting ... before refresh date.What reloads on every refresh, such as 7 days.
Detect data changes.Reloads a period only if an audit column's max changed. Not the partition column; misses hard deletes.
Only refresh complete days.Skips today's partial day.
Get the latest data in real time with DirectQuery.A hybrid table; Premium-type workspaces; related tables in Dual.

Partitions appear on the first refresh in the service. Never republish a partitioned model from Desktop; deploy metadata only through XMLA. Direct Lake doesn't use incremental refresh.

Performance fixes (L26)

Symptom.Fix.
Performance analyzer shows a big DAX query time.Tune the measure or the model.
Big Other time on many visuals.Fewer visuals per page.
A large model.Drop unused columns and rows, lower cardinality, turn off auto date/time.
FILTER(table) inside CALCULATE.A Boolean column filter, with KEEPFILTERS if needed.
IFERROR around a division.DIVIDE.
The same expression written twice.Store it once in a VAR.

[236][245][250][251][252][257][258]

You are an analytics engineer at Juniper Coffee. The Sales table's incremental refresh filter is [OrderDate] >= RangeStart and [OrderDate] <= RangeEnd. Finance reports that some days' totals are doubled. What should you change?

Pick one answer.

[250]

Maintain: access, security and governance

Exam depth: know it cold.

Workspace roles (L27): the lowest role that can

Task.Lowest role.
View reports, query Lakehouse or Warehouse with T-SQL.Viewer.
Create, edit and delete items, build on models, read with Spark.Contributor.
Share items, publish an app, add people at Member or lower.Member.
Delete the workspace, remove anyone, connect Git.Admin.

Item and model permissions (L27)

One person needs.Grant.
One item only.Share that item, not a workspace role.
T-SQL reads.ReadData (Read all with SQL analytics endpoint).
Spark reads.ReadAll (Read all with Apache Spark).
Reports, composite models or Analyze in Excel on a model.Build.
Republish, Refresh now, XMLA changes.Write, which also bypasses RLS.
Share onward.Reshare.

The highest role from any group wins. Read alone connects but queries nothing. Read without Build doesn't secure data; RLS does. Many read-only consumers get an app with audiences. Grant to groups. A revoked permission can take up to two hours to apply.

[265][267][271][272][273][274]

Which security layer (L28)

Who reads the data.Secure it in.
Report users only.The semantic model: RLS roles with DAX, OLS.
T-SQL users of a Warehouse or SQL endpoint.Warehouse T-SQL: GRANT, column-level security, RLS policy, masking.
Spark, SQL and Direct Lake users of a Lakehouse, or its files.OneLake security roles.

Dynamic RLS compares to USERPRINCIPALNAME() through a security table. A user in two roles sees the union. Anyone with Write on the model bypasses RLS and OLS: Admins, Members and Contributors. OLS hides tables and columns, set in TMDL view or Tabular Editor. Add role members in the service.

Label, badge or security (L29)

You need to.Use.
Flag how sensitive an item is.Sensitivity label, from Microsoft Purview.
Say it's ready to share, as its creator.Promoted: anyone with write permission.
Mark that it passed an official review.Certified: admin-authorized reviewers.
Mark the one reference copy of core data.Master data: admin-authorized users, data items only.
Restrict who sees data.Roles, RLS or OLS.

Applying a label needs Pro or PPU and edit permission. Inheritance never overwrites a manual label or swaps in a weaker one, and stops past 80 downstream items. Dashboards can't be endorsed.

[151][209][247][273][278][284][287][290]

You are an analytics engineer at Juniper Coffee. A finance analyst must build her own reports on the sales semantic model. She must not change the model or see other items in its workspace. What follows least privilege?

Pick one answer.

[265][273]

Maintain: lifecycle and reusable assets

Exam depth: know it cold.

Git or deployment pipelines (L30, L31)

You need.Use.
History, review, rollback, merging work.Git integration.
Promote content from Dev to Test to Prod.A deployment pipeline.
Different data sources or parameters per stage.Deployment rules on the target stage.

The usual pattern: only Dev connects to Git, and a pipeline promotes Dev to Test to Prod.

  • Git: an Admin connects the workspace. Contributors commit and update. Commit picks items and is blocked while updates wait; Update takes the whole branch. A .pbip stores the model as TMDL and the report as PBIR; .gitignore leaves out cache.abf.
  • Pipelines: 2 to 10 stages. A deploy copies metadata, never data, permissions, RLS members, refresh schedules or credentials. Rules are created on the target stage by the item's owner. Deploying needs pipeline admin plus Contributor on both workspaces.
  • Impact analysis crosses workspaces and needs write permission. Lineage view stays inside one workspace.

XMLA endpoint (L32): Read Only by default. A capacity admin sets Read Write. Writes need it: Tabular Editor edits, partition refresh, metadata-only deploys.

Which reusable asset (L32)

You need to reuse.Use.
The data and logic, for many reports.A shared semantic model, with Build.
The design, with new data each time.A .pbit template, which holds no data.
Just the connection to a source.A .pbids file, with no credentials.
A shared model plus your own tables.A composite model.

[246][294][306][308][310][312][318]

Facts to know cold

Exam depth: know it cold. Read these the night before.

  • Roles: Viewer has ReadData, not ReadAll or Build. Contributor is the lowest role that creates items. Only an Admin connects Git.
  • Permissions: Write on a model bypasses RLS. Build is what report authors need. Read alone queries nothing.
  • SQL analytics endpoint: read-only T-SQL, Delta tables only.
  • Storage modes: Import refreshes up to 48 times a day on Fabric capacity. Direct Lake needs a Fabric capacity, refreshes by framing, and on OneLake never falls back.
  • Fallback: DirectLakeBehavior defaults to Automatic. SQL RLS or a SQL view makes Direct Lake on SQL fall back.
  • Model size: 1 GB in the small format; large format is required past 10 GB.
  • Refresh: >= RangeStart and < RangeEnd. Partitions appear on the first refresh in the service.
  • DAX: SUM adds a column, SUMX adds per row. ALLSELECTED respects slicers, ALL ignores them.
  • Defaults: Fabric stopped creating a default semantic model on September 5, 2025. A merge with no join kind is left outer.
  • Mirroring: replication compute is free, plus storage up to 1 TB per CU.
  • OneLake integration: exports Import tables only, not measures or DirectQuery tables.
  • Deployment pipelines: 2 to 10 stages; deploys copy metadata, not data.

[4][12][40][50][89][198][265][305]

Exam day: question types, the clock and the constraint word

Exam depth: know it cold.

The format. DP-600 gives you 100 minutes, with about 120 in the seat. Most Microsoft exams have 40 to 60 questions. A wrong answer costs nothing, so answer every one. Microsoft doesn't publish each exam's question types, but its exam sandbox (aka.ms/examdemo) lets you try them all:

  • Multiple choice: one answer, or several when the question says how many.
  • Drag and drop, and build list: move pieces into slots or into order.
  • Hot area: click part of a picture, or pick from dropdowns inside code.
  • Case study: a company described across tabs, then several questions about it.
  • Yes/no series: one goal, a different proposed solution each time. Judge each on its own; several can work, or none. Candidates report you can't go back once you answer, and Microsoft allows no break inside one.

The clock. Keep 20 to 30 minutes for the case study, as candidates advise. Answer every question on the first pass. If unsure, pick your best guess and mark it for review. Check the clock at the halfway question. Use the review screen last. A break doesn't stop the clock, and afterwards you can't return to earlier questions.

Microsoft Learn opens in a split screen, with no extra time. Ctrl+F finds text on a page. Microsoft says that if you look up every question, you won't finish. Save it for one or two exact facts, such as a setting name.

The constraint word. Read the last sentence first. Find the phrase that decides it: least privilege, no data movement, without a refresh, no code, smallest model. Cross out options that break it, then pick.

[1][3][4][320]

Three answers would work. The constraint word picks one. Find it before you read the options. Picture: Vintage instruction-manual drawing, cream paper, fine black line art: step 1 shows a cartoon figure reading the last line of an exam questio

The night before. Read the tables and facts on this page, and do nothing new. Testing online? Run the system test on the same computer, network and room you'll use. Check that the name on your ID matches your certification profile exactly. Then sleep.

The morning of. Run the system test again. Online, you need a walled room with the door closed and nobody else in it. Clear the desk, unplug extra monitors, and keep your phone out of reach but within hearing. Check-in opens 30 minutes before your appointment and closes 15 minutes after. It takes about 15 minutes, so start early.

Online or test center. A test center gives you a quiet, ready-made setup. Online is almost anywhere, but your room and computer must pass the checks. Both need one original, unexpired, government-issued ID with your name, photo and signature.

If it doesn't go well. Wait 24 hours for the second try, then 14 days between later tries, at most five in 12 months. Retakes are paid where a fee applies. Once you pass, the certification lasts one year, and renewal is a free online assessment in the six months before it expires.

[5][7][321][322][323]

Exam-style practice

A yes/no series asks how to stop Juniper's store managers seeing other regions' rows in a report. Solution one, semantic model RLS with USERPRINCIPALNAME(), looked right, so you answered Yes. Solution two proposes the same rule with a static role per region. What should you do?

Pick one answer.

[3][320]

You are an analytics engineer at Juniper Coffee. A matrix shows stores on rows and a Rank measure. A manager picks three cafés in a Store slicer. Those three must rank 1, 2, 3 among themselves. Which measure works?

Pick one answer.

[226][227]

FlashcardsIndex cardsFlashcards ⚡

Card 1 of 19

[3][45][87][198][250][265][273][323]

Worth keeping

  1. Every choose-between question has a deciding phrase; each table here pairs one with its answer.
  2. Biggest point swings: Direct Lake flavors and fallback, storage modes, least-privilege permissions and RLS.
  3. 100 minutes, no penalty for guessing: answer everything, mark, and keep 20-30 minutes for the case study.
  4. In a yes/no series, judge each solution alone and answer once.
  5. Book the exam when you score 80% or more on both mocks.

One test left before you book. Mock exam 2 has 50 new questions on the same blueprint and a different case study. Sit it in one go, with the timer, and see whether your weakest area from mock 1 has moved past 80%.

Leaderboard

Loading…

50 XP for each lesson you finish, 10 for each right answer. Names show only for learners who turn on Show me on leaderboards.

Sources for this lesson (92)

Each fact in the lesson is marked with the number of the page it comes from.

  1. [1]Microsoft Learn: Microsoft Certified: Fabric Analytics Engineer Associate
  2. [3]Microsoft Learn: Exam duration and exam experience
  3. [4]Microsoft Learn: Exam scoring and score reports
  4. [5]Microsoft Learn: Renew your Microsoft Certification
  5. [6]Microsoft Learn: Study guide for Exam DP-600: Implementing Analytics Solutions Using Microsoft Fabric
  6. [7]Microsoft Learn: Register and schedule an exam
  7. [12]Microsoft Learn: What is a lakehouse in Microsoft Fabric?
  8. [34]Microsoft Learn: Dimensional modeling in Microsoft Fabric Warehouse: Dimension tables
  9. [40]Microsoft Learn: What is Mirroring in Fabric?
  10. [45]Microsoft Learn: Direct Lake overview
  11. [50]Microsoft Learn: Power BI semantic models in Microsoft Fabric
  12. [60]Microsoft Learn: Model relationships in Power BI Desktop
  13. [68]Microsoft Learn: Workspace identity
  14. [70]Microsoft Learn: What is an on-premises data gateway?
  15. [71]Microsoft Learn: What is a virtual network (VNet) data gateway?
  16. [81]Microsoft Learn: Microsoft Fabric decision guide: choose a data store
  17. [82]Microsoft Learn: Microsoft Fabric decision guide: Choose between Warehouse and Lakehouse
  18. [83]Microsoft Learn: OneLake shortcuts
  19. [84]Microsoft Learn: OneLake shortcut security
  20. [87]Microsoft Learn: Microsoft Fabric decision guide: Choose a data movement strategy
  21. [88]Microsoft Learn: Turn on OneLake availability for an eventhouse
  22. [89]Microsoft Learn: OneLake integration for semantic models
  23. [90]Microsoft Learn: Microsoft Fabric decision guide: copy activity, Copy job, dataflow, Eventstream, or Spark
  24. [92]Microsoft Learn: Dataflow Gen2 data destinations and managed settings
  25. [94]Microsoft Learn: What is Copy job in Data Factory for Microsoft Fabric?
  26. [96]Microsoft Learn: Ingest data into your Warehouse using Transact-SQL
  27. [98]Microsoft Learn: Using the data profiling tools
  28. [103]Microsoft Learn: Merge queries overview
  29. [104]Microsoft Learn: Append queries
  30. [113]Microsoft Learn: Table.NestedJoin
  31. [121]Microsoft Learn: Working with duplicate values
  32. [123]Microsoft Learn: Analytical and aggregate functions in Fabric Data Warehouse
  33. [125]Microsoft Learn: TRY_CAST (Transact-SQL)
  34. [139]Microsoft Learn: T-SQL surface area in Fabric Data Warehouse
  35. [142]Microsoft Learn: ALTER VIEW (Transact-SQL)
  36. [143]Microsoft Learn: CREATE FUNCTION (Fabric Data Warehouse)
  37. [151]Microsoft Learn: Secure your Fabric Data Warehouse
  38. [156]Microsoft Learn: IDENTITY columns in Fabric Data Warehouse
  39. [168]Microsoft Learn: Query using the visual query editor
  40. [175]Microsoft Learn: KQL quick reference
  41. [197]Microsoft Learn: CALCULATE
  42. [198]Microsoft Learn: How Direct Lake works
  43. [199]Microsoft Learn: Develop Direct Lake semantic models
  44. [201]Microsoft Learn: Semantic model modes in the Power BI service
  45. [203]Microsoft Learn: What is Power BI Premium?
  46. [204]Microsoft Learn: Large semantic models in Power BI Premium
  47. [205]Microsoft Learn: Many-to-many relationship guidance
  48. [206]Microsoft Learn: Bi-directional relationship guidance
  49. [207]Microsoft Learn: Active vs inactive relationship guidance
  50. [209]Microsoft Learn: Row-level security (RLS) with Power BI
  51. [214]Microsoft Learn: DIVIDE function vs. divide operator (/)
  52. [224]Microsoft Learn: SUMX
  53. [226]Microsoft Learn: RANKX
  54. [227]Microsoft Learn: ALLSELECTED
  55. [229]Microsoft Learn: WINDOW
  56. [232]Microsoft Learn: ISINSCOPE
  57. [236]Microsoft Learn: Appropriate use of error functions
  58. [237]Microsoft Learn: Calculation groups
  59. [239]Microsoft Learn: Use field parameters to let report readers change visuals
  60. [240]Microsoft Learn: Use composite models in Power BI
  61. [242]Microsoft Learn: User-defined aggregations
  62. [245]Microsoft Learn: Configure incremental refresh and real-time data for Power BI semantic models
  63. [246]Microsoft Learn: Semantic model connectivity with the XMLA endpoint
  64. [247]Microsoft Learn: Integrate Direct Lake security
  65. [250]Microsoft Learn: Configure incremental refresh and real-time data
  66. [251]Microsoft Learn: Advanced incremental refresh and real-time data with the XMLA endpoint
  67. [252]Microsoft Learn: Use Performance Analyzer to examine report performance
  68. [257]Microsoft Learn: Data reduction techniques for Import modeling
  69. [258]Microsoft Learn: Auto date/time guidance in Power BI Desktop
  70. [265]Microsoft Learn: Roles in workspaces in Microsoft Fabric
  71. [267]Microsoft Learn: How lakehouse sharing works
  72. [271]Microsoft Learn: Roles in workspaces in Power BI
  73. [272]Microsoft Learn: Publish a workspace app in Power BI
  74. [273]Microsoft Learn: Semantic model permissions
  75. [274]Microsoft Learn: Share items in Microsoft Fabric
  76. [278]Microsoft Learn: Table, column, and row-level security in OneLake
  77. [284]Microsoft Learn: Apply sensitivity labels to Fabric items
  78. [287]Microsoft Learn: Endorsement overview
  79. [290]Microsoft Learn: Information protection in Microsoft Fabric
  80. [294]Microsoft Learn: Basic concepts in Git integration
  81. [305]Microsoft Learn: Introduction to deployment pipelines
  82. [306]Microsoft Learn: The deployment pipelines process
  83. [308]Microsoft Learn: Create deployment rules
  84. [310]Microsoft Learn: Choose the best Fabric CI/CD workflow option for you
  85. [311]Microsoft Learn: Lineage in Fabric
  86. [312]Microsoft Learn: Impact analysis
  87. [315]Microsoft Learn: Power BI usage scenarios: Advanced data model management
  88. [318]Microsoft Learn: Create and use report templates in Power BI Desktop
  89. [320]Datakuity: Tips for the DP-700 exam (candidate report)
  90. [321]Microsoft Learn: About online exams with Pearson VUE
  91. [322]Microsoft Learn: Exam and assessment lab security policy
  92. [323]Microsoft Learn: Exam and Applied Skills assessment lab retake policies