Black Friday50% off your first month on any monthly plan with EARLY50 See plans
All 33 lessons
  1. Start here: cloud and data 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. The three languages: SQL, PySpark and KQL side by side
    6. 6. Getting data in: the map of Fabric tools
  2. Store, ingest and transform batch data (exam area 2, 30-35%)

    1. 7. Choosing a data store and a transformation tool
    2. 8. Shortcuts and mirroring: use data without copying it
    3. 9. Ingest data with pipelines and Copy job
    4. 10. Transform with PySpark and SQL: join, denormalize, aggregate
    5. 11. Messy data: duplicates, missing values and late arrivals
    6. 12. Full and incremental loads, and loading a star schema
  3. Orchestrate and stream (exam areas 1 and 2)

    1. 13. Pipeline or notebook? Choosing how to orchestrate
    2. 14. Schedules and event-based triggers
    3. 15. Parameters and dynamic expressions
    4. 16. Streaming in Fabric: Eventstreams, Spark streaming and KQL
    5. 17. KQL and windowing functions for real-time data
  4. Set up and secure a Fabric solution (exam area 1, 30-35%)

    1. 18. Workspace settings: Spark, domains, OneLake and data workflows
    2. 19. Version control with Git and database projects
    3. 20. Deployment pipelines: development, test and production
    4. 21. Who can see what: workspace roles and item permissions
    5. 22. Fine-grained security: rows, columns, objects, files and masking
    6. 23. Governance: sensitivity labels and endorsement
  5. Monitor, fix and optimize (exam area 3, 30-35%)

    1. 24. Monitoring hub, alerts and Activator
    2. 25. Fixing errors in pipelines, dataflows and notebooks
    3. 26. Fixing errors in Eventhouse, Eventstream and T-SQL
    4. 27. Optimize lakehouse tables: V-Order, OPTIMIZE, VACUUM and partitions
    5. 28. Optimize Spark, pipelines and warehouses
    6. 29. Optimize eventstreams, eventhouses and queries
  6. Exam practice and final review

    1. 30. Exam day: question types, timing and how to think
    2. 31. Mock exam 1
    3. 32. Mock exam 2
    4. 33. Final review: every decision on one page
See the leaderboard

18 min

Data basics: files, tables, formats and the lakehouse idea

Tell files from tables, pick the right format, and see why Fabric stores tables as Delta.

By the end of this lesson you can

  • Choose between CSV, JSON, Parquet and Delta for a given need
  • Explain what Delta Lake adds to Parquet: ACID and time travel
  • Tell a data warehouse, a data lake and a lakehouse apart
  • Sort workloads into OLTP or OLAP, and batch or streaming

From earlier lessons · 1 of 3

Which cloud model is Microsoft Fabric?

Pick one answer.

From earlier lessons · 2 of 3

What does the number in an F SKU, such as F64, tell you?

Pick one answer.

From earlier lessons · 3 of 3

Where is a workspace's data stored?

Pick one answer.

Northwind Rides' first data has arrived. The repair partner sent a folder of CSV files. The scooters send JSON events every few seconds. Payments sit in SQL Server tables. Your manager asks: "Can't we just keep it all as CSV? Everyone can open those."

[4]

Structured, semi-structured and unstructured data

Exam depth: recognize it.

  • Structured data follows a fixed schema, a set list of columns and their types. Every row has the same fields. Northwind's payments table is structured.
  • Semi-structured data has some structure, but fields can vary from record to record. A scooter event in JSON (JavaScript Object Notation) may carry a battery field that another event lacks.
  • Unstructured data has no set structure: photos of damaged scooters, PDF repair reports, audio.

Fabric stores all three. Tables hold structured data. Files hold the rest until you shape them into tables.

[29]

File formats: CSV, JSON, Parquet and Delta

Exam depth: know it cold.

FormatHow it stores dataGood for
CSVPlain text, one row per line, commas between fields.Swapping simple tables between any tools.
JSONPlain text with nested fields that can vary.Events and app data, such as scooter messages.
ParquetCompressed binary, columnar: each column's values sit together.Fast analytics on large data.
DeltaParquet files plus a transaction log of every change.Reliable tables you update. Fabric's default.

CSV and JSON are easy for people to read but slow to scan at scale. Parquet is built for machines.

Decision rule: receive data in whatever format it comes in, then store it for analysis as Delta tables.

[11][12]

Predict, then reveal

Northwind's trips table has 60 columns and 500 million rows. The ops team asks for the average trip length. Predict why this runs much faster on Parquet than on CSV.

[11]

Delta Lake: Parquet plus a transaction log

Exam depth: know it cold.

Delta Lake is an open table format. A Delta table is a folder of Parquet files plus a transaction log, a record of every change made to the table.

The log gives you two things plain Parquet can't.

ACID transactions. Each change is all-or-nothing and leaves the data valid:

  • Atomic: a change succeeds completely or not at all.
  • Consistent: data moves from one valid state to another.
  • Isolated: two jobs writing at once don't see each other's half-done work.
  • Durable: once saved, a change stays saved.

Time travel. Every saved change creates a new table version. You can query the table as it was at an earlier version or time.

When you load data into Fabric, it's stored as Delta tables by default. Every Fabric engine reads them.

[12][13][32][34]

sql
-- See which versions exist
DESCRIBE HISTORY trips;
-- Read the table as it was at version 5
SELECT * FROM trips VERSION AS OF 5;
-- Read the table as it was on a date
SELECT * FROM trips TIMESTAMP AS OF '2026-11-01';
Time travel in Spark SQL, as you'd run it in a Fabric Notebook, a page for running code. Reading an old version doesn't change the table.

[34]

Delta keeps every version. Time travel is reading an old page, not rewinding the table. Picture: Claymation-style scene: a small clay librarian character flips back through a thick clay ledger book labeled "TRANSACTION LOG".

[34]

You are a data engineer at Northwind Rides. A faulty job ran this morning and overwrote the Delta table of station data with wrong values. Before you fix it, you need to see exactly what the table held yesterday. You must not change the table yet. What should you use?

Pick one answer.

[34]

Data warehouse, data lake, lakehouse

Exam depth: know it cold.

StoreWhat it isStrength
Data warehouseTables in a relational schema, queried with SQL.Fast, trusted reporting on structured data.
Data lakeCheap storage for huge numbers of files in any format.Holds everything, including raw and unstructured data.
LakehouseA data lake with tables on top of its files.Lake storage with warehouse-style queries.

A relational schema is a set of tables linked by key columns.

Fabric has an item for each idea. A Lakehouse holds files and Delta tables, used by Spark and SQL. A Warehouse gives full T-SQL, Microsoft's dialect of SQL, including writes. You'll meet both properly in lesson 4.

[21][30]

Rendering diagram…
A lakehouse keeps the lake's files and adds tables you can query like a warehouse.

[30]

The lakehouse: keep every file like a lake, query it like a warehouse. Picture: Whiteboard doodle in black and blue marker.

OLTP vs OLAP

Exam depth: recognize it.

OLTPOLAP
Stands forOnline transaction processing.Online analytical processing.
WorkMany small reads and writes as things happen.Big reads that summarize history.
ExampleRecording a payment when a trip ends.Revenue by station by month.
Optimized forFast writes that keep data correct.Fast queries over lots of rows.

Northwind's SQL Server database is OLTP: it runs the rental app. Your Fabric platform is the OLAP side. You copy data out of the OLTP system and reshape it for analysis, so reports don't slow the app down.

[13][30]

Batch vs streaming

Exam depth: know it cold.

Batch processing collects records and processes them together, for example every night. Stream processing handles each record as it arrives.

BatchStreaming
Data scopeThe whole dataset.Recent records, or a rolling time window.
LatencyUsually hours.Seconds or less.
Typical workComplex analysis.Simple counts, alerts and rolling averages.

Latency is the delay between data arriving and the result being ready.

Decision rule: if the need says near real-time or as it happens, think streaming. If a nightly or hourly result is fine, batch is simpler.

Many platforms do both, and Northwind will too.

[31]

Match

Match each Northwind Rides workload to the kind of processing it is

Pick an item on the left, then its match on the right.

[13][31]

A first look at the star schema

Exam depth: recognize it. Lesson 12 teaches loading one in full.

A star schema is the standard table design for analysis. It has two kinds of table:

  • A fact table stores measurements of events: each trip, with its duration, distance and fare. Those number columns are measures.
  • Dimension tables describe the things involved: stations, customers, dates. Their text columns are what you filter and group by.

The fact table sits in the middle and the dimensions around it, like the points of a star.

Each dimension row gets a surrogate key, a single-column ID the warehouse generates. The ID from the source system is the natural key. Surrogate keys protect you when sources change or clash.

[35][36][37]

Rendering diagram…
Northwind's trips as a star: one fact table, three dimensions joined by surrogate keys.

[35]

Exam-style practice

You are a data engineer at Northwind Rides. Trip data must be queried fast by Spark and SQL. Nightly jobs update rows in it, and two jobs sometimes write at the same moment. Readers must never see half-written data. You need to choose the storage format. What should you use?

Pick one answer.

[11][12]

You are a data engineer at Northwind Rides. The ops team wants a screen that shows how many scooters are moving in each district, updated within seconds. You need to choose how to process the trip events. What should you use?

Pick one answer.

[31]

FlashcardsIndex cardsFlashcards ⚡

Card 1 of 15

[11][12][13][29][30][31][34][35]

Worth keeping

  1. CSV and JSON are for moving data; Parquet is columnar and fast for analytics.
  2. Delta = Parquet + a transaction log, giving ACID and time travel. It's Fabric's default.
  3. A lakehouse is a lake with tables on top; a warehouse is relational tables with SQL.
  4. OLTP runs the app; OLAP answers questions. Streaming means seconds; batch means hours.
  5. A star schema has a fact table of measures and dimension tables with surrogate keys.

Files, Delta tables, a lake, a warehouse, streaming events: that sounds like five separate systems to run. In Fabric it's one. Next lesson you open OneLake, the single lake every Fabric item writes to.

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 (16)

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

  1. [4]Microsoft Learn: Study Guide for Exam DP-700
  2. [10]Microsoft Learn: Understand Microsoft Fabric licenses and capacity
  3. [11]Microsoft Learn: Explore file storage
  4. [12]Microsoft Learn: Delta Lake overview
  5. [13]Microsoft Learn: Explore transactional data processing
  6. [15]Microsoft Learn: Configure Multi-Geo support for Fabric
  7. [17]Microsoft Learn: Shared responsibility in the cloud
  8. [21]Microsoft Learn: Fabric terminology
  9. [29]Microsoft Learn: Identify data formats
  10. [30]Microsoft Learn: Explore analytical data processing
  11. [31]Microsoft Learn: Understand batch and stream processing
  12. [32]Microsoft Learn: Delta Lake table format interoperability
  13. [34]Microsoft Learn: Query Delta tables with time travel
  14. [35]Microsoft Learn: Dimensional modeling in Fabric Data Warehouse
  15. [36]Microsoft Learn: Dimensional modeling in Microsoft Fabric Warehouse: Dimension tables
  16. [37]Microsoft Learn: Dimensional modeling in Microsoft Fabric Warehouse: Fact tables