- Secure, durable, scalable, and easily accessible from across the globe.
- E.g. persistent data across devices for mobile applications.
- Uses REST API endpoints that make data available to huge range of application types & platforms e.g. .NET, JAVA, NODE.
- Automated backup and recovery: mitigates the risk of losing data if there is any unforeseen failure or interruption.
- Replication across the globe
- Copies your data to protect it against any planned or unplanned events
- e.g. scheduled maintenance or hardware failures.
- Allows you to replicate your data at multiple locations across the globe.
- Copies your data to protect it against any planned or unplanned events
- Support for data analytics: supports performing analytics on your data consumption.
- Encryption capabilities: You have tight control over who can access the data.
- Multiple data types: Almost any e.g. videos, text, like binary files. Many options for SQL and NoSQL data.
- Data storage in virtual disks: Up to 32 TB. Significant when you're storing heavy data such as videos and simulations.
- Storage tiers: To prioritize access to data based on frequently used vs rarely used information.
- Also called relational data
- Data that adheres to a schema.
- Defines table, fields, clear relationship between two
- Can be stored in e.g. database table with rows and columns.
- Relies on keys to indicate how one row in a table relates to data in another row of another table.
- 💡 It's easy to enter, query, and analyze.
- All of the data follows the same format.
- E.g. sensor data or financial data.
- Azure services:
- Azure SQL Database
- Azure Cosmos DB (SQL API)
- Also called as non-relational or NoSQL data.
- Doesn't fit neatly into tables, rows, and columns.
- Instead uses tags or keys that organize and provide a hierarchy for the data.
- Azure services:
- Azure Cosmos DB (MongoDB API, Cassandra API)
- Azure Table Storage
- Azure Queue Storage
- Encompasses data that has no designated structure to it
- There are no restrictions on the kinds of data it can hold.
- e.g. PDF document, a JPG image, a JSON file, video content, etc
- 💡More prominent as businesses try to tap into new data sources.
- Azure services:
- Azure Blob Storage
- Azure File Storage
- Azure Data Lake Storage
- Azure Disk Storage
- Includes disks attached to virtual machines, file shares, databases
- They can expand & shrink necessarily
- Common characteristics:
- Durable and highly available with redundancy and replication.
- Secure through automatic encryption and role-based access control.
- Scalable with virtually unlimited storage.
- Managed, handling maintenance and any critical problems for you.
- Accessible from anywhere in the world over HTTP or HTTPS.
- Also known as Azure blobs
- Good for very large objects, such as video files or bitmaps
- Unstructured, meaning that there are no restrictions on the kinds of data it can hold.
- Can manage thousands of simultaneous uploads, massive amounts of video data, constantly growing log files, and can be reached from anywhere with an internet connection.
- Lets you
- Stream large video or audio files directly to the user's browser from anywhere in the world.
- Send large volumes of data directly to the browser.
- 💡 Also used to store data for backup, disaster recovery, and archiving.
- 📝 Ability to store up to 8 TB of data for virtual machines (VM disks)
- Hot storage tier: optimized for storing data that is accessed frequently.
- Cool storage tier: optimized for data that are infrequently accessed and stored for at least 30 days.
- Archive storage tier: for data that are rarely accessed and stored for at least 180 days with flexible latency requirements.
- Also known as Azure disks
- Provides disks for virtual machines, applications, and other services to access and use as they need.
- In the background they are page-blobs in a blob storage
- Allows data to be persistently stored and accessed from an attached virtual hard disk.
- Disks can be managed or unmanaged by Azure, and therefore managed and configured by the user.
- 💡 Use-case examples: Lift and shift
- Storing data that is not required to be accessed from outside the virtual machine to which the disk is attached.
- Different sizes and performance levels
- Solid-state drives (SSDs)
- Hard disk drives (HDDs)
- 💡 Use standard SSD and HDD disks for less critical workloads
- Premium SSD disks for mission-critical production applications.
- Durable: ZERO% annualized failure rate.
- 📝 Allows you to perform analytics on your data usage and prepare reports.
- Stores both structured and unstructured data.
- Combines the scalability and cost benefits of object storage with the reliability and performance of the Big Data file system capabilities.
- Supports batch queries, interactive queries, real-time analytics, machine learning, and being a data warehouse.
- Also known as Azure files
- File shares that you can access and manage like a file server
- Fully managed file shares in the cloud that are accessible via the industry standard Server Message Block (SMB) protocol.
- Ensures the data is encrypted at rest and in transit.
- Can be mounted concurrently by cloud or on-premises Windows, Linux, and macOS.
- 📝 Any number of Azure virtual machines or roles can mount and access the file storage share simultaneously.
- 💡 Good to share files anywhere in the world, diagnostic data, or application data sharing.
- A data store for queuing and reliably delivering messages between applications
- 💡 Helps build flexible applications and separate functions for better durability across large workloads
- When application components are decoupled, they can scale independently
- Provides asynchronous message queueing for communication between application components
- Typically, there are one or more sender components and one or more receiver components.
- Sender components add messages to the queue
- Receiver components retrieve messages from the front of the queue for processing
- 💡 Use-case examples:
- Create a backlog of work and to pass messages between different Azure web servers.
- Distribute load among different web servers/infrastructure and to manage bursts of traffic.
- Build resilience against component failure when multiple users access your data at the same time.
- NoSQL data store
- Schema-less design
- For data at rest helps you secure your data.
- It encrypts the data before storing it and decrypts the data before returning it.
- Encryption & decryption are transparent to the user.
- Data is already encrypted by the client libraries.
- Azure stores the data in the encrypted state at rest, which is then decrypted during retrieval.
- Set up when you create a storage account
- Ensures that your data is durable and always available
- Provides regional and geographic replications
- Protects data against natural disasters and other local disasters like fire or flooding.
- Cost effectiveness
- Pay-as-you go
- No dedicated hardware to be purchased, installed, configured and maintained. = no up-front expense (or capital cost).
- Scalable: No need to have idle hardware
- Reliability
- Managed data backup, load balancing, disaster recovery, and data replication as services to ensure data safety and high availability.
- Storage types
- On-premises => often requires numerous servers and administrative tools for each storage type.
- Azure has different storage options for each part of your solution.
- Agility
- Requirements and technologies change: No need to reprovisioning & deployment of new infrastructure.
- Create new services in minutes = change storage back-ends quickly without needing a significant hardware investment.
| Needs | On-premises | Azure data storage |
|---|---|---|
| Compliance and security | Dedicated servers required for privacy and security | Client-side encryption and encryption at rest |
| Store structured and unstructured data | Additional IT resources with dedicated servers required | Azure Data Lake and portal analyzes and manages all types of data |
| Replication and high availability | More resources, licensing, and servers required | Built-in replication and redundancy features available |
| Application sharing and access to shared resources | File sharing requires additional administration resources | File sharing options available without additional license |
| Relational data storage | Needs a database server with database admin role | Offers database-as-a-service options |
| Distributed storage and data access | Expensive storage, networking, and compute resources needed | Azure Cosmos DB provides distributed access |
| Messaging and load balancing | Hardware redundancy impacts budget and resources | Azure Queue provides effective load balancing |
| Tiered storage | Management of tiered storage needs technology and labor skill set | Azure offers automated tiered storage of data |