Are you upgrading from SQL Server 2005, 2008, or 2008R2?

APPLIES TO: yesSQL Server (Windows only) noAzure SQL Database noAzure SQL Data Warehouse noParallel Data Warehouse

The end of extended support for SQL Server is one reason to upgrade now to a newer version of SQL Server and to Azure SQL Database. Upgrading enables you to maintain security and compliance, achieve breakthrough performance, and optimize your data platform infrastructure.

For more info, guidance, and tools to plan and automate your upgrade or migration, see SQL Server 2005 end of support and SQL Server 2008 end of support.

Why upgrade?

Important

Extended support for SQL Server 2005 ended on April 12, 2016. If you're still running SQL Server 2005 after April 12, 2016, you no longer receive security updates.

Important

Extended support for SQL Server 2008 and 2008r2 ended on July 09, 2019. If you're still running SQL Server 2008 or 2008R2 after July 09, 2019, you will no longer receive security updates. More information can be found in the blog Announcing new options for SQL Server 2008. To extend support for free, you can migrate your SQL Server to an Azure VM. For more information see Extend support for SQL Server 2008 and 2008 R2 with Azure.

Choose your upgrade option

If you're upgrading relational databases from an older version of SQL Server, here are your options for relational storage on the Microsoft platform.

To see a more comprehensive analysis of these options, see PaaS vs IaaS.

Relational storage option Benefits Other factors to consider
SQL Server hosted on Azure virtual machines

Consider this option if you want the following:

Benefits of migrating to a hosted environment.

Control over the operating environment.

Familiar feature set of SQL Server.
You can extend support for SQL Server 2008 and 2008 R2 for free for up to 3 years.

You can deploy quickly from a library of virtual machine images.

You get the full SQL Server feature set.

You save the cost of hardware and of server software. You pay only for hourly usage.
You have to manage both SQL Server and the operating system software.



For more information, see SQL Server on Azure Virtual Machines overview.

For info about migrating, see Migrate a database to SQL Server on an Azure VM.
Azure SQL Database managed instance (PaaS)

Consider this option if you want a lower-cost solution with less maintenance.

A managed instance is similar to an instance of the Microsoft SQL Server database engine, offering shared resources for databases and additional instance-scoped features.

Managed instance supports database migration from on-premises with minimal to no database change.
Get the benefit of cross-database queries within the same managed instance, as well as CLR and SQL job support.

99.995% availability guaranteed.

The cost of the service includes not only storage, but high availability, patching and automated backups.
There are some Transact-SQL (T-SQL) differences between Azure SQL Database managed instance and SQL Server on-premises. For more info, see Azure SQL Database managed instance T-SQL information.

For more information about SQL Database managed instance, see Azure SQL Database managed instance overview and Azure SQL Database managed instance capabilities.

For information about migrating, see Migrating a SQL Server to Azure SQL Database managed instance.
Azure SQL Database single database or elastic pool (PaaS)

Consider this option if you want a lower-cost solution with less maintenance.

This option is particularly well suited for cloud-designed applications when developer productivity and fast time-to-market for new solutions are critical, or that have to provide external access.

The most commonly used SQL Server features are available, but not as many as for an Azure SQL Database managed instance.
You can deploy quickly and scale up easily.

99.995% availability guaranteed.

You can pay for usage by the second or the hour.

The cost of the service includes not only storage, but high availability patching, and automated backups.
There are some Transact-SQL (T-SQL) differences between Azure SQL Database and SQL Server on-premises. For more info, see Azure SQL Database Transact-SQL information.

Azure SQL Database also has a maximum database size of 100 TB, compared to 524 PB for SQL Server. For more information, see Resource limits for single databases

For more information about SQL Database, see Azure SQL Database overview and Azure SQL Database documentation.

For information about migrating, see Migrating a SQL Server database to Azure SQL Database.
SQL Server on premises

Consider this option for database applications of any kind, from transactional systems to data warehouses.
You have the most control over features and scalability because you manage both hardware and software.

If you're upgrading from an older instance of SQL Server, this is the most similar environment.
You have to make the biggest up-front investment and provide the most ongoing management, because you have to buy, maintain, and manage your own hardware and software.

For more information, see SQL Server.

You may also want to consider a non-relational or NoSQL solution for certain data and applications.

Non-relational solution Benefits
Azure Cosmos DB

Consider this option for modern, scalable, mobile, and web applications that use JSON data and require a combination of robust querying and transactional data processing.

For more info, see Cosmos DB.

For info about importing data, see Import data to Cosmos DB.
Your documents are indexed and you can use familiar SQL syntax to query them.

The database is schema-free.

You can add properties to documents without having to rebuild indexes.

You get JSON and JavaScript support right inside the database engine.

You get native support for geospatial data and integration with other Azure Services including Azure Search, HDInsight, and Data Factory.

You get low latency, high-performance storage with reserved throughput levels.
Azure table storage

Consider this option to store petabytes of semi-structured data in a cost-effective solution.

For more info, see Table Storage.
You can evolve your apps and your table schema without taking the data offline.

You can scale up without sharding your dataset.

You get geo-redundant storage that replicates data across multiple regions.

Plan your upgrade

Get SQL Server

To download an evaluation copy of SQL Server, see SQL Server downloads.

Next Steps

SQL Server 2017
SQL Server 2005 end of support
SQL Server 2008 end of support