Azure SQL Database,

Azure SQL Managed Instance

or
Michelle Gutzait
SQL Server VMs. MS Data Platform Architect at
www.Pythian.com
gutzait@Pythian.com

Why, when, what ?

Are you already in “the Cloud”?
Migration to the Cloud – easy or complex?
Where’s your data?
Real Hybrid

Hybrid cloud
On-premises & cloud
Consistent experience is everything
Microsoft Worldwide DCs – end of 2017
Advantages of Microsoft Azure

• Fast deployment of Applications

• Disaster recovery and High Availability
• Security
• Manageability - Built-in support for patching, log
analytics, monitoring and site recovery.
• Flexibility – Dynamic use of new services and
scale data storage and machine capabilities.
• Overall costs
 PaaS

IaaS
Cloud Option

• IaaS – Azure VMs

• PaaS – Azure Databases
• Managed Instances
• Azure SQL Database
(“singleton”)
• Azure SQL Database Elastic
Pool
IaaS - SQL Server on Virtual Machines
• Easily “lift-and-shift migrations
• Set up and manage your SQL Server deployments in minutes, and
then scale up or down (based on potential transition rules)
• Highly-secured and trusted
• Work with any SQL Server version and edition on any virtual
machine size
• Employ virtual machines that are tuned for exceptional
performance by the team that builds SQL Server
• Pay per minute, or bring your own licenses
• 3 tiers: Basic, Low Priority, Standard
PaaS - Azure SQL Database Deployment
Options

● Near 100% ● Own set of ● Collection of databases

compatibility resources ● Shared set of resources
with on- managed via a ● Managed via a logical
premises SQL logical server server
Server instances
Single databases can be moved into and out
of the Elastic Pool. No accessible system
databases
Single SQL Database vs. Elastic pool

• A singleton database - dedicated resources when

having a detailed workload pattern.
• Elastic Pool supports Multiple Single databases with
unpredictable varying workloads. Typically:
• 3 or more databases
• Varying and unpredictable resource consumption (CPU,
memory, and IO).
DTU-based vs. vCore-based purchasing
models
• DTU-based purchasing model
• Offers a blend of compute, memory, and IO resources in three service
tiers:
• Basic,
• Standard,
• Premium.
• Compute sizes within each tier provide a different mix of these resources,
to which you can add additional storage resources.

• vCore-based purchasing model

• Lets you choose the number of vCores, the amount of memory, and the
amount and speed of storage.
Managed Instance is available in two service tiers
• General Purpose: Designed for applications with typical performance and IO latency
requirements.
• Design for the majority of business applications with typical performance requirements
• High-performance Azure Premium storage (8 TB)
• Built-in High-availability based on reliable Azure Premium Storage and Azure Service Fabric

• Business Critical: Designed for applications with low IO latency requirements and minimal
impact of underlying maintenance operations on the workload.
• Designed for business applications with highest performance and HA requirements
• Comes with super-fast SSD storage (up to 1 TB on Gen 4 and up to 4 TB on Gen 5)
• Built-in High availability based on Always On Availability Groups and Azure Service Fabric.
• Built-in additional Read-only database replica that can be used for reporting and other read-only workloads
• In-Memory OLTP that can be used for workload with high-performance requirements

Both service tiers guarantee 99.99% availability and enable you to

independently select storage size and compute capacity.
Automatically switching between the tiers is not supported at the moment.
Migration is required in that case.
Azure High Availability
• Basic, Standard, and General Purpose service tier availability
• Replicated Storage
• Premium and Business Critical service tier availability
• Always On Availability Groups with replicated storage
• Premium
• Availability Group
• Business Critical
• Availability Group with Read-Only replica
• Zone redundant configuration
• Same Data Centre by default
• Can be changed to multiple locations
• Service Fabric
• data (for example, in the Basic HA, there is a stateless compute layer that is running the sqlserver.exe process and
contains only transient and cached ample – plan cache, buffer pool, column store pool). This stateless SQL Server node
is operated by Azure Service Fabric that initializes process, controls health of the node, and performs failover to another
place if necessary.
• In addition, there is a stateful Storage layer for the database files guaranteeing no data loss.

• https://docs.microsoft.com/en-us/azure/sql-database/sql-database-high-
availability#premium-and-business-critical-service-tier-availability
Additional topics not discussed here…

• Network
• Security
• Application Stack
• Identity Management (i.e.: Citrix)
• How to migrate
• …
Let’s Compare!

Here are some guidelines…

Microsoft
Guidelines
• Type of systems to move
and where to
• Licensing and Pricing
models
• How to calculate overall
costs
• SLAs
• Where to start

• https://docs.microsoft.com/e
n-us/azure/sql-database/sql-
database-paas-vs-sql-server-
iaas
https://docs.microsoft.com/en-us/azure/sql-database/sql-database-features
Main differences
SQL Server on VM
• You have full control over the SQL •
Server engine
• Up to 99.95% availability
• Full parity with the matching version •
of on-premises SQL Server
• Fixed, well-known database engine
version
• Previous engine versions support
• Easy migration from SQL Server on-
premises
• Private IP address within Azure Vnet
• You have ability to deploy
application or services on the host
where SQL Server is placed.
Managed instance in SQL Database
High compatibility with SQL Server
on-premises
• 99.99% availability guaranteed
Built-in backups, patching, recovery
• Latest stable Database Engine
version (lower comp. level
supported)
• Easy migration from SQL Server.
Private IP address within Azure Vnet
• Built-in advanced intelligence and
security
• Online change of resources
(CPU/storage)
• Automatic tuning available
Single database / elastic pool in
SQL Database
• The most commonly used SQL
Server features are available
• 99.99% availability guaranteed
• Built-in backups, patching, recovery
• Latest stable Database Engine
version (lower comp. level
supported)
• Ability to assign necessary resources
(CPU/storage) to individual
databases
• Built-in advanced intelligence and
security
• Online change of resources
(CPU/storage).
• Automatic tuning available
Comparison of Limitations
SQL Server on VM Managed instance in SQL Database Single database / elastic pool in
SQL Database
• SQL Server instances with • Up to 8 TB storage • Databases of up to 100 TB
up to 64 TB of storage

• 32,767 databases per • Up to 100 databases per • Up to 5000 databases per

instance instance server

• Supports Cross-database • Supports Cross-database • Does not support Cross-

queries queries database queries
Price Comparison
• Hardware and Infrastructure
(purchase, renewing)
• Technical Team
• Recurring Costs (such as:
electricity)
• Backup and recovery
• High Availability and Disaster
Recovery
• Security
• Choose features
• Usage per month
• Can use resource pools
• Can shut down for more saving
Price calculator:
https://azure.microsoft.com/en-
ca/pricing/calculator/?&OCID=AID71
9803_SEM_oLvc2Eyv&lnkd=Google_
Azure_Brand&dclid=CNfD7OX84N4
CFWawswodafIABw
DEMO
DTU Calculator tool allows you to upload CSV file with
following metrics (https://dtucalculator.azurewebsites.net/)
WHERE TO
Data Migration Assistant (DMA)
DMA replaces all previous versions of SQL Server Upgrade Advisor and
should be used for upgrades for most SQL Server versions. Supported source
and target versions are:
• Sources
• SQL Server 2005
• SQL Server 2008
• SQL Server 2008 R2
• SQL Server 2012
• SQL Server 2014
• SQL Server 2016
• SQL Server 2017 on Windows
• Targets
• SQL Server 2012
• SQL Server 2014
• SQL Server 2016
• SQL Server 2017 on Windows and Linux
• Azure SQL Database
• Azure SQL Database Managed Instance
Next step
Open Discussion and QA
Thank
you!