🔒 Guided

Pre-launch preview. Authorised access only.

Incorrect code

Guided by A Guide to Cloud
Explore AB-900 AI-901
Guided DP-900 Domain 1
Domain 1 — Module 4 of 7 57%
4 of 27 overall

DP-900 Study Guide

Domain 1: Core Data Concepts

  • Your First Look at Data Free
  • Data File Formats: CSV, JSON, Parquet & More Free
  • Databases: Relational vs Non-Relational Free
  • Transactional Workloads: Keeping Data Consistent Free
  • Analytical Workloads: Finding the Insights Free
  • Data Roles: DBA, Engineer & Analyst Free
  • The Azure Data Landscape Free

Domain 2: Relational Data on Azure

  • Relational Data: Tables, Keys & Relationships
  • Normalization: Why Duplicate Data is Bad
  • SQL Basics: SELECT, INSERT, UPDATE, DELETE
  • Database Objects: Views, Indexes & More
  • Azure SQL: Your Database in the Cloud
  • Open-Source Databases on Azure
  • Choosing the Right Azure Database

Domain 3: Non-Relational Data on Azure

  • Azure Blob Storage: Files in the Cloud
  • Azure Files & Table Storage
  • Azure Cosmos DB: The Global Database
  • Cosmos DB APIs: SQL, MongoDB & More
  • Choosing Non-Relational Storage

Domain 4: Analytics on Azure

  • Data Ingestion & Processing
  • Analytical Data Stores: Data Lakes, Warehouses & Lakehouses
  • Microsoft Fabric & Azure Databricks
  • Batch vs Streaming: Two Speeds of Data
  • Real-Time Analytics on Azure
  • Power BI: See Your Data
  • Data Models in Power BI
  • Choosing the Right Visualization

DP-900 Study Guide

Domain 1: Core Data Concepts

  • Your First Look at Data Free
  • Data File Formats: CSV, JSON, Parquet & More Free
  • Databases: Relational vs Non-Relational Free
  • Transactional Workloads: Keeping Data Consistent Free
  • Analytical Workloads: Finding the Insights Free
  • Data Roles: DBA, Engineer & Analyst Free
  • The Azure Data Landscape Free

Domain 2: Relational Data on Azure

  • Relational Data: Tables, Keys & Relationships
  • Normalization: Why Duplicate Data is Bad
  • SQL Basics: SELECT, INSERT, UPDATE, DELETE
  • Database Objects: Views, Indexes & More
  • Azure SQL: Your Database in the Cloud
  • Open-Source Databases on Azure
  • Choosing the Right Azure Database

Domain 3: Non-Relational Data on Azure

  • Azure Blob Storage: Files in the Cloud
  • Azure Files & Table Storage
  • Azure Cosmos DB: The Global Database
  • Cosmos DB APIs: SQL, MongoDB & More
  • Choosing Non-Relational Storage

Domain 4: Analytics on Azure

  • Data Ingestion & Processing
  • Analytical Data Stores: Data Lakes, Warehouses & Lakehouses
  • Microsoft Fabric & Azure Databricks
  • Batch vs Streaming: Two Speeds of Data
  • Real-Time Analytics on Azure
  • Power BI: See Your Data
  • Data Models in Power BI
  • Choosing the Right Visualization
Domain 1: Core Data Concepts Free ⏱ ~10 min read

Transactional Workloads: Keeping Data Consistent

When money moves, inventory changes, or orders are placed, every step must succeed — or none of them should. That's the world of transactional workloads.

What is a transactional workload?

☕ Simple explanation

A transactional workload is any system where data must be 100% accurate, right now.

Think of paying for coffee with your debit card. Three things happen: your bank subtracts $4.00, the café’s bank adds $4.00, and a receipt is created. If the café’s bank fails to add the money, your bank should NOT subtract it either. All steps succeed together, or none of them happen. That’s a transaction.

Systems that handle these kinds of “all-or-nothing” operations are called transactional workloads — also known as OLTP (Online Transaction Processing).

A transactional workload (OLTP — Online Transaction Processing) handles high volumes of small, discrete operations that create, read, update, or delete individual records. These operations must maintain data consistency at all times, even when thousands of users access the system simultaneously.

Transactional systems enforce the ACID properties to guarantee that every transaction either completes fully or has no effect. They are the backbone of operational systems — point-of-sale, order management, banking, inventory tracking, and booking systems.

ACID: the four guarantees

Every transactional system follows ACID properties — four rules that ensure data stays correct:

PropertyWhat It MeansTom’s Delivery Example
AtomicityAll steps in a transaction succeed, or none of them doA delivery is marked “complete” AND the driver’s route is updated — or neither happens
ConsistencyData always moves from one valid state to anotherThe total number of packages in the system always matches what was loaded onto trucks
IsolationConcurrent transactions don’t interfere with each otherTwo dispatchers assigning drivers at the same time don’t accidentally assign the same driver twice
DurabilityOnce a transaction is committed, it’s permanent — even if the power goes outOnce Tom confirms a delivery, that record survives a server restart
ℹ️ ACID analogy: the bank transfer

The classic example: transferring $100 from Account A to Account B.

  1. Atomicity: Subtract $100 from A AND add $100 to B — both steps or neither
  2. Consistency: Total money in the system stays the same (A + B = constant)
  3. Isolation: Another person transferring money at the same time sees consistent balances
  4. Durability: Once the transfer is confirmed, it’s saved permanently

If the system crashes between step 1 (subtract from A) and step 2 (add to B), atomicity ensures the subtraction is rolled back. The money doesn’t disappear.

Characteristics of transactional workloads

FeatureDescription
High volume, small operationsThousands of inserts, updates, and deletes per second
Low latencyEach operation completes in milliseconds
Current dataAlways shows the latest state (not historical trends)
Multiple concurrent usersMany users reading and writing at the same time
ACID complianceEvery operation follows the four ACID properties
Normalised schemaData is organised to minimise duplication (normalisation)

Real-world examples:

  • Priya’s FreshMart: point-of-sale system recording every item scanned at checkout
  • Tom’s Pacific Freight: dispatch system assigning drivers to deliveries
  • Aisha’s food app: order placement (choose items → confirm → pay)
  • Jake’s CloudPulse: user account creation, subscription management

OLTP vs the next module

Transactional workloads (OLTP) are about recording what’s happening right now. The next module covers analytical workloads (OLAP), which are about understanding what happened in the past to make better decisions.

Transactional vs analytical workloads — a preview
FeatureTransactional (OLTP)Analytical (OLAP)
PurposeRecord current operationsAnalyse historical data
OperationsInsert, update, delete individual recordsRead large volumes, aggregate, summarise
DataCurrent state (live data)Historical snapshots (over time)
UsersMany concurrent users (clerks, apps, customers)Fewer users (analysts, managers)
Speed priorityFast individual operations (milliseconds)Fast bulk queries (seconds to minutes)
SchemaNormalised (minimise duplication)Denormalised (optimise for reading)
💡 Exam tip: OLTP keyword spotting

When the exam describes a scenario, look for these OLTP signals:

  • “Recording sales as they happen” → OLTP
  • “Multiple cashiers entering data simultaneously” → OLTP
  • “Must ensure no duplicate bookings” → OLTP (isolation)
  • “If payment fails, the order should be cancelled” → OLTP (atomicity)

If the scenario talks about “trends,” “reports,” “dashboards,” or “historical analysis” — that’s analytical (next module).

Flashcards

Question

What does ACID stand for?

Click or press Enter to reveal answer

Answer

Atomicity (all or nothing), Consistency (valid state to valid state), Isolation (concurrent transactions don't interfere), Durability (committed data survives failures).

Click to flip back

Question

What does OLTP stand for, and what does it do?

Click or press Enter to reveal answer

Answer

Online Transaction Processing. OLTP systems handle high volumes of small read/write operations — recording current business activities like sales, orders, and bookings.

Click to flip back

Question

Why is atomicity important in transactional systems?

Click or press Enter to reveal answer

Answer

Atomicity ensures that all steps of a transaction succeed together or fail together. Without it, a crash mid-transaction could leave data in an inconsistent state (e.g., money deducted but not credited).

Click to flip back

Knowledge check

Knowledge Check

Tom's dispatch system needs to assign a driver to a delivery. The system must: (1) mark the driver as 'busy,' (2) link the driver to the delivery, and (3) update the delivery status. If step 2 fails, what should happen?

Knowledge Check

Which ACID property ensures that two customers buying the last concert ticket at the same time don't BOTH get it?

🎬 Video coming soon

Next up: Analytical Workloads: Finding the Insights — from recording data to understanding it.

← Previous

Databases: Relational vs Non-Relational

Next →

Analytical Workloads: Finding the Insights

Guided

I learn, I simplify, I share.

A Guide to Cloud YouTube Feedback

© 2026 Sutheesh. All rights reserved.

Guided is an independent study resource and is not affiliated with, endorsed by, or officially connected to Microsoft. Microsoft, Azure, and related trademarks are property of Microsoft Corporation. Always verify information against Microsoft Learn.