Choosing the Right Data Store--An Overview of Azure and Fabric Data Platform Choices

#fabcon2026 #session #learning #microsoft #fabcon2026-session

Description

Choosing the right data store can make or break the performance and costs of your application. In this session, we will cover all the options available to you in the Azure and Fabric Data space. You'll learn about CosmosDB, Azure/Fabric SQL DB, Postgres, and MySQL services, as well as Databricks.

Key Takeaways

Denny Cherry & Associates Consulting
Microsoft Certified Master
Allows for specific versions of SQL Server to be deployed
Allows for a specific build of SQL Server to be used for as long as
Running SQL VMs allows you to control the OS and SQL Patching
Allows for storage size only limited by the size of the VM
Allows for different data files or databases to be placed on

My Notes

Action Items

Resources & Links

Slides

Choosing the Right
Data Store
SQL VM, SQL DB, Managed
Instance, Cosmos,
Synapse, Hadoop
About Me
• Denny Cherry & Associates Consulting
• 8 books
• Dozens of articles
• Microsoft MVP
• Microsoft Certified Master
• VMware vExpert
Agenda
SQL VM
SQL DB
Managed Instance
Cosmos
Synapse
Platform Services
Security &
Management
Compute
Cloud
Services
Developer Services
Web and Mobile
Service
Fabric
API
Management
API
Apps
Web Apps
Hybrid
Operations
Azure SDK
Visual Studio
Azure AD
Connect Health
Portal
Batch
Active
Directory
Multi-Factor
Authentication
Automation
Hybrid
Connections
Key Vault
Store /
Marketplace
Biztalk
Services
Service
Bus
Media & CDN
Media
Services
VM Image Gallery
& VM Depot
Notification
Hubs
Content Delivery
Network (CDN)
Application
Insights
Team Project
Data
Analytics & IoT
Integration
Storage
Queues
Logic
Apps
Mobile
Apps
Remote App
AD Privileged
Identity
Management
Backup
HDInsight
Machine
Learning
Azure
SQL DB
Data
Factory
Event
Hubs
Redis
Cache
Search
Stream
Analytics
Mobile
Engagement
Cosmos
Tables
Synapse
Operational
Insights
Import/Export
Site
Recovery
StorSimple
Infrastructure Services
Compute
Virtual
Machines
Storage
Containers
BLOB Storage
Azure Files
Networking
Premium
Storage
Virtual
Network
Load
Balancer
Datacenter Infrastructure (55+ Regions)
DNS
Express
Route
Traffic
Manager
VPN
Gateway
Application
Gateway
SQL VMs
SQL VMs
Gives you full control over the SQL Server Instance and the VM OS
Gives you full control of the HA and DR configuration to meet your
requirements
Allows for specific versions of SQL Server to be deployed
Allows for a specific build of SQL Server to be used for as long as
needed by the application
SQL VMs
Running SQL VMs allows you to control the OS and SQL Patching
schedules
Gives fine grained control of firewall rules allowing specific servers
access
Allows for storage size only limited by the size of the VM
Allows for different data files or databases to be placed on
different storage configurations
SQL VMs
SQL is used for relational systems, either OLTP workloads or data
warehouse workloads
Databases can be configured across multiple disk volumes allowing for
maximum IO throughput
Everything you expect from SQL Server works, as it is just SQL Server
Monolithic databases can be created without issue
Reserved Instances are available for most VM sizes
Azure
SQL
SQL DBs
Four flavors of SQL DB
Azure SQL – SQL Server Based
Azure Database for MySQL
Azure Database for PostgreSQL
Azure Database for MariaDB
Azure vs. Fabric
Azure SQL DB and Fabric SQL DB are effectively the same
Same database engine for both
Fabric SQL DB supports a subset of features that Azure SQL DB
supports
Azure SQL
Cloud hosted database
Cloud database platform
Databases can be configured as a single database or within an
elastic pool
Applications can be migrated from on-prem to Azure SQL DB but
the most successful migrations will involve application redesign
Azure SQL
By scaling horizontally across Azure an application can have
limitless scale
Single databases are scaled independently from the other
databases
Elastic Pools are a group of databases which share resources
single a single pool of CPU and memory resources
Azure SQL
SQL DBs are available as DTU or vCore
vCore is available in two models
Gen 4 – Physical Cores – Only for legacy deployments
Gen 5 – Threaded Cores
FSv2 Series
M-Series
Azure SQL
Gen 4 vCores
Limited to 24 Cores per database / pool
Azure SQL
Gen 5 Cores
Limited to 128 cores and 625 GB of RAM
Based on a variety of CPUs
Intel E5-2673 v4 (Broadwell) 2.3 GHz
Intel SP8160 (Skylake) processors
Intel Xeon Platinum 8272CL 2.5 GHz (Cascade Lake) processors.
Azure SQL
Hyperscale databases are available
Up to 128 TB of storage
Available in all Azure regions
Supports Serverless
Reservations are not available for HyperScale
Azure SQL
FSv2 Series
Limited to 72 vcores and 136 GB of RAM
Based on Intel Xeon® Platinum 8168 (SkyLake) processors
Hyperscale databases are available
Available in a lot of Azure regions including: Australia Central, Australia Central 2, Australia East,
Australia Southeast, Brazil South, Canada Central, East Asia, East US, France Central, India
Central, Korea Central, Korea South, North Europe, South Africa North, Southeast Asia, UK
South, UK West, West Europe, West US 2
Supports Serverless
Azure SQL
Perfect for OLTP databases
Can be used to Data warehouses depending on performance and
size needs
Best use involves a scale out configuration of the database tier
Azure SQL DB can be mirrored into Fabric
Open Source Databases
More like Managed Instance than SQL DB
Hosted as a MySQL/PostgreSQL/MariaDB instance
Managed
Instances
Managed Instance
SQL Server Instance in the cloud
Solves the issues that come from Azure SQL DB
Cross database queries
MSDB / SQL Server Agent
Linked Servers
Managed Instance
Up to 128 Cores
Read replica available on Business Critical Tier
Can be mirrored into Fabric
Managed Instance
Limited to 500 database per Managed Instance
Limited to 32 TB per Managed Instance
Great for OLTP databases that can’t be scaled out in SQL DB, or
applications that need a feature not supported by SQL DB
Cosmos
Cosmos
Autoscale platform
Several different engines and APIs within the cosmos platform
Can be configured within multiple regions
Can be configured as multi-master
Data is partitioned for maximum performance
Cosmos
Gremlen API
Cassandra API
SQL API
MongoDB API
Table API
Cosmos
Infinite scale of the data store
Billing is priced in RUs
The more data that is returned to more RUs that are used
Large data sets can be very expensive
Reserved instances are available
Cosmos
Ideal for logging applications
Great for application with small consistent query workloads
Synapse
Synapse
Scale out MPP Database platform
Great solution for data warehouses
Requires careful database design to correctly scale data across the
partitions
Limited concurrent queries, no not a good viable for OLTP
workloads
Synapse
Scales across multiple nodes on the back end
Databases can be scaled wider or narrower as needed extremely
quickly
Storage is allocated in 1TB chunks
Storage is effectively unlimited
Reserved Instances are available
Synapse in Fabric
Synapse is available as part of Fabric as Datawarehouse
Fabric Data Warehouse is the next generation of data warehousing in
Microsoft Fabric
Fabric Data Engineering is our big data analytics workload in Fabric
Synapse Data Science empowers data scientists to explore their data, build
and operationalize their predictive models
Synapse Real-time Analytics is a robust platform tailored to deliver realtime data insights and observability analytics capabilities for a wide range
of data types
Questions?
Sound off.
The mic is all yours.
Influence the product roadmap.
Join the Fabric User Panel
Join the SQL User Panel
Share your feedback directly with our
Fabric product group and researchers.
Influence our SQL roadmap and ensure
it meets your real-life needs
https://aka.ms/JoinFabricUserPanel
https://aka.ms/JoinSQLUserPanel