Chapter 10 of 20
Azure Storage Services: Blobs, Files, Disks, and Data Redundancy
Dive into Azure’s core storage offerings and redundancy options so you can pick the right storage type and durability level for any scenario.
Big Picture: Azure Storage in the AZ-900 World
Storage as the Third Pillar
Alongside compute and networking, Azure Storage is a core pillar. It provides cloud-based data services hosted in Microsoft data centers that you access over the internet or private links.
Where Storage Fits in the Cloud
Azure Storage underpins both IaaS and PaaS. You do not manage physical disks; you consume services like blobs, file shares, queues, and tables, usually via a storage account.
Key AZ-900 Expectations
For AZ-900, you must recognize when to use blobs, file shares, queues, tables, and managed disks, and compare redundancy options like LRS, ZRS, GRS, and GZRS for durability and availability.
Storage Accounts: The Foundation of Azure Storage
What Is a Storage Account?
A storage account is a secure, configurable container that holds multiple data services like blobs, file shares, queues, and tables, all under one unique name and region.
Core Properties
Each storage account has a unique name, a region, a performance tier (Standard or Premium), and an account kind, usually general-purpose v2 for modern features.
Services Inside a Storage Account
In a GPv2 account you can create blob containers, file shares, queues, and tables. The storage account acts as the main boundary for configuration and billing.
Security Integration
Storage accounts integrate with Microsoft Entra ID and RBAC so you can control which users or apps can manage the account and, in many cases, access data inside it.
Blob Storage: Object Storage for Unstructured Data
What Is Blob Storage?
Blob Storage is Azure's object storage for unstructured data like images, videos, documents, and backups. Data is stored as blobs inside containers in a storage account.
Containers and Blobs
Think of a storage account as a building, containers as folders, and blobs as files. You organize blobs into containers for management and security.
Blob Types
Block blobs store general files and media, append blobs are for log-like data you only add to, and page blobs are used mainly for Azure managed disks.
Access Tiers
Hot is for frequently accessed data, cool for infrequent access, and archive for long-term storage with high retrieval latency but very low cost.
Typical Blob Scenarios
Use blobs for app uploads, media libraries, analytics data lakes, and backups. Exam phrases like 'large unstructured data' usually point to Blob Storage.
Azure Files: Cloud File Shares with SMB
What Is Azure Files?
Azure Files provides managed file shares in the cloud that you access over SMB or NFS. To apps and users, it looks like a traditional network drive.
How Azure Files Is Used
You create file shares inside a storage account and mount them on Windows, Linux, or macOS, or access them via REST APIs for scripts and apps.
Common Scenarios
Typical uses include replacing on-prem file servers, storing shared app configs, or hybrid setups with Azure File Sync caching data on-premises.
Exam Comparison: Files vs Blobs
Pick Azure Files for SMB-based shared folders and lift-and-shift file servers. Pick Blob Storage for object storage accessed mainly via HTTP or APIs.
Performance Tiers
Standard file shares are HDD-backed and cheaper. Premium file shares are SSD-backed and suited for IO-intensive enterprise workloads.
Queues and Tables: Messaging and NoSQL Storage
Queue Storage Basics
Azure Queue Storage holds many small messages. Producers add messages, and workers remove and process them, enabling asynchronous, decoupled processing.
When to Use Queues
Use queues to decouple app components, handle background jobs, and smooth out spikes in workload by buffering work items.
Table Storage Basics
Azure Table Storage is a NoSQL key-attribute store. Data is stored as entities with PartitionKey and RowKey plus flexible properties.
When to Use Tables
Use tables for large volumes of semi-structured data like logs or device data where you do not need complex relational queries.
Exam Tip: Tables vs Cosmos DB
Table Storage is simple and cheap inside a storage account. Azure Cosmos DB is a more advanced, globally distributed NoSQL database service.
Managed Disks: Storage for Azure Virtual Machines
What Are Managed Disks?
Managed disks are Azure-managed virtual disks for VMs. They use page blobs under the hood, but Azure hides storage account details from you.
Disk Roles
Managed disks attach to VMs as OS disks or data disks. You choose size and performance; Azure manages placement, scaling, and availability.
Disk Types
Standard HDD is cheapest, Standard SSD balances cost and performance, and Premium SSD offers high performance and low latency for production.
Redundancy for Disks
Managed disks are LRS by default. For higher resilience, you can use availability zones and zone-aware options in your VM architecture.
Exam Scenario Cues
High IO production VM? Choose Premium SSD. Dev or test with low IO? Standard disks are usually sufficient and more cost-effective.
Redundancy Basics: Availability vs Durability
Why Redundancy Matters
Redundancy means storing multiple copies of your data so that hardware, rack, or even regional failures do not cause data loss or long outages.
Availability vs Durability
Availability is how often your data is reachable. Durability is how likely it is to remain intact over time despite failures. Replication boosts both.
Replication Trade-offs
More replication across zones or regions improves resilience but increases cost. AZ-900 expects you to match redundancy level to business needs.
Scenario Keywords
Look for phrases like 'data center failure', 'regional outage', or 'lowest cost' to infer which redundancy option fits best.
Redundancy Options: LRS, ZRS, GRS, RA-GRS, GZRS, RA-GZRS
LRS: Local Redundancy
LRS keeps three copies of your data in one data center within a region. It protects against disk and server failures, and is the lowest-cost option.
ZRS: Zone Redundancy
ZRS spreads three copies across different availability zones in a region, protecting against data center failures while keeping data in one region.
GRS and RA-GRS
GRS adds a second region: data is LRS in the primary and replicated to a paired region. RA-GRS lets you read from that secondary region.
GZRS and RA-GZRS
GZRS combines ZRS in the primary region with geo-replication to a secondary region. RA-GZRS adds read access to the secondary for DR and reporting.
Choosing Options on the Exam
Cheapest and basic protection? LRS. Protect from data center failure? ZRS. Protect from regional outage? GRS or GZRS, with RA variants for read access.
Matching Storage Types and Redundancy to Real Scenarios
Scenario 1: Static Website Assets
Global site stores product images and scripts, infrequent updates, cached by CDN. Use Blob Storage with hot tier and LRS in the origin region.
Scenario 2: Replace File Server
Apps access a shared path over SMB; data must stay in-country. Use Azure Files plus ZRS to keep data in one region but survive a data center failure.
Scenario 3: Financial Records
Critical transaction logs must survive regional disasters and support analytics reads. Use blobs or tables with RA-GZRS for geo and zone resilience.
Reading Scenario Cues
Look for keywords like 'SMB', 'static website', or 'regional outage' to infer which storage type and redundancy option the question is pointing toward.
Thought Exercise: Design a Storage Strategy
Work through this design mentally (or jot notes) to solidify your understanding.
Company background
You are advising a small online education startup (perfect timing) that runs:
- A web app where students upload assignment PDFs.
- A set of Windows VMs running a legacy reporting tool that expects a shared drive.
- A background job that processes uploaded files and writes status messages.
- A requirement to keep data available even if a single data center in the region fails, but they are not ready to pay for a secondary region yet.
Your task
For each need, choose:
- The Azure storage service: Blob, Files, Queue, Table, or managed disks.
- An appropriate redundancy option: LRS or ZRS (they do not want geo yet).
Try to answer before reading the suggested solution below.
Pause here and decide...
Suggested design
- Student uploads: Blob Storage (block blobs) in a GPv2 storage account, ZRS for zone resilience.
- Legacy reporting tool: Azure Files SMB share, ZRS so the share survives a data center failure.
- Background job status messages: Azure Queue Storage for work items; possibly Table Storage for status history, both with ZRS.
- VM disks: Managed disks using at least Standard SSD, deployed across availability zones for higher resilience.
Reflect:
- Where did you pick the same answers? Where did you differ, and why?
- How would your choices change if the startup said "cheapest possible" or "must survive regional disasters" instead?
Quiz 1: Pick the Right Storage Service
Test your ability to choose between blobs, files, queues, tables, and managed disks.
A company needs to store large numbers of high-resolution product photos and make them available over HTTPS to a global web application. The photos are read frequently, but updated rarely. Which Azure storage option is the best fit?
- Azure Files with SMB shares
- Azure Blob Storage using block blobs
- Azure Queue Storage
- Azure Table Storage
Show Answer
Answer: B) Azure Blob Storage using block blobs
Block blobs in Azure Blob Storage are designed for storing large unstructured objects like images and serving them over HTTP or HTTPS. Azure Files is for SMB file shares, Queue Storage is for messages, and Table Storage is for NoSQL key-attribute data.
Quiz 2: Choose the Correct Redundancy Level
Now check your understanding of redundancy options.
Your organization stores critical documents in Azure Storage and must remain operational if an entire data center in the selected region goes offline. Regulations require data to stay within that region, and management does not want to pay for geo-replication. Which redundancy option is most appropriate?
- Locally redundant storage (LRS)
- Zone-redundant storage (ZRS)
- Geo-redundant storage (GRS)
- Read-access geo-redundant storage (RA-GRS)
Show Answer
Answer: B) Zone-redundant storage (ZRS)
Zone-redundant storage (ZRS) replicates data across multiple availability zones within a single region. It protects against data center failures while keeping data in the same region. GRS and RA-GRS replicate to a paired region, which the scenario explicitly avoids. LRS only protects within one data center.
Flashcards: Core Azure Storage Concepts
Flip these cards mentally to reinforce key terms before you move on to practice questions and mock exams.
- Storage account
- A secure, scalable container in Azure that provides a unique namespace and configuration boundary for services like Blob Storage, Azure Files, Queues, and Tables.
- Blob Storage
- Azure's object storage service for unstructured data such as images, videos, documents, and backups, organized as blobs within containers in a storage account.
- Azure Files
- A managed file share service in Azure that exposes file shares over SMB or NFS, allowing you to replace or extend traditional file servers.
- Azure Queue Storage
- A simple message queue service used to store and retrieve messages for asynchronous, decoupled communication between application components.
- Azure Table Storage
- A NoSQL key-attribute store for large volumes of semi-structured data, using tables, entities, PartitionKey, and RowKey.
- Managed disk
- An Azure-managed virtual disk resource used by virtual machines, built on top of page blobs, where Azure handles storage accounts and replication.
- Locally redundant storage (LRS)
- A redundancy option that stores three copies of your data within a single data center in one Azure region, protecting against drive and server failures.
- Zone-redundant storage (ZRS)
- A redundancy option that stores multiple copies of your data across different availability zones in a region, protecting against data center failures within that region.
- Geo-redundant storage (GRS)
- A redundancy option that combines LRS in the primary region with asynchronous replication to a paired secondary region, protecting against regional outages.
- Read-access geo-redundant storage (RA-GRS)
- Like GRS but with read access to the secondary region, enabling read-only workloads or disaster recovery access from the secondary.
- Geo-zone-redundant storage (GZRS) and RA-GZRS
- Redundancy options that combine ZRS in the primary region with geo-replication to a secondary region; RA-GZRS also allows read access to the secondary.
Key Terms
- Azure Files
- A managed file share service in Azure that exposes file shares over SMB or NFS, allowing you to replace or extend traditional file servers.
- Blob Storage
- Azure's object storage service for unstructured data such as images, videos, documents, and backups, organized as blobs within containers in a storage account.
- Managed disk
- An Azure-managed virtual disk resource used by virtual machines, built on top of page blobs, where Azure handles storage accounts and replication.
- Storage account
- A secure, scalable container in Azure that provides a unique namespace and configuration boundary for services like Blob Storage, Azure Files, Queues, and Tables.
- Azure Queue Storage
- A simple message queue service used to store and retrieve messages for asynchronous, decoupled communication between application components.
- Azure Table Storage
- A NoSQL key-attribute store for large volumes of semi-structured data, using tables, entities, PartitionKey, and RowKey.
- Geo-redundant storage (GRS)
- A redundancy option that combines LRS in the primary region with asynchronous replication to a paired secondary region, protecting against regional outages.
- Zone-redundant storage (ZRS)
- A redundancy option that stores multiple copies of your data across different availability zones in a region, protecting against data center failures within that region.
- Locally redundant storage (LRS)
- A redundancy option that stores three copies of your data within a single data center in one Azure region, protecting against drive and server failures.
- Geo-zone-redundant storage (GZRS)
- A redundancy option that combines ZRS in the primary region with geo-replication to a secondary region, improving resilience to both zone and regional failures.
- Read-access geo-redundant storage (RA-GRS)
- Like GRS but with read access to the secondary region, enabling read-only workloads or disaster recovery access from the secondary.
- Read-access geo-zone-redundant storage (RA-GZRS)
- A redundancy option that combines GZRS-style replication with read access to the secondary region for disaster recovery and reporting.