View a markdown version of this page

Scenario 1: Microsoft SQL Server Migration to AWS - Microsoft Workloads Lens - AWS Well-Architected Framework
CharacteristicsReference architectureConfiguration notes

Scenario 1: Microsoft SQL Server Migration to AWS

Organizations running business-critical SQL Server workloads often face challenges with hardware lifecycle management, licensing costs, and operational overhead. Migrating SQL Server to AWS offers opportunities for improved scalability, reduced maintenance burden, and potential cost savings through various deployment options. This scenario covers approaches from lift-and-shift to complete modernization, helping organizations choose the right path based on their specific requirements.

Characteristics

  • Organizations are dealing with SQL Server workloads that are essential for business operations.

  • Current challenges include managing hardware lifecycles, SQL Server licensing expenses, and operational maintenance.

  • The scenario spans different migration approaches, from basic lift-and-shift to full modernization.

  • The focus is on helping organizations select migration strategies based on their unique needs.

Reference architecture

Reference architecture diagram

  1. Self-Managed SQL Server on EC2: This approach transfers SQL Server workloads from on-premises servers to self-managed EC2 instances using native SQL Server tools or full server replication methods. It's ideal for organizations needing detailed control or preserving complex configurations. Migration options include database-level methods (for example, backup and restore, log shipping) or server-level replication using tools like AWS Application Migration Service.

  2. Self-Managed SQL Server on EC2 (AWS DMS): This approach migrates SQL Server workloads from on-premises servers to self-managed EC2 instances using AWS Database Migration Service. DMS facilitates data migration and synchronization, offering a streamlined process for transferring database contents while minimizing downtime. It's suitable for organizations seeking a managed migration tool with replication capabilities.

  3. Amazon RDS for SQL Server (Windows Methods): This strategy replatforms SQL Server workloads from on-premises servers to Amazon RDS using native SQL Server migration tools. Methods like backup/restore or log shipping are employed to transfer databases to the managed RDS environment, balancing familiar SQL Server native capabilities with the benefits of a managed service.

  4. Amazon RDS for SQL Server (AWS DMS): This approach replatforms SQL Server workloads to Amazon RDS using AWS Database Migration Service. DMS manages data migration and synchronization from on-premises servers to RDS, combining the advantages of a managed migration process with the operational benefits of RDS for ongoing database management.

  5. Modernize to Amazon Aurora: This path transforms SQL Server workloads to Amazon Aurora, leveraging AWS Schema Conversion Tool for schema conversion to PostgreSQL and AWS DMS for data migration. For applicable use cases, Babelfish for Aurora PostgreSQL can be employed to minimize application code changes, offering a balance between modernization and compatibility.

Configuration notes

  • Migrating to self-managed SQL Server on Amazon EC2 is often an effective strategy for projects with tight deadlines. This approach allows for a quick lift-and-shift migration while maintaining full control over the database environment. Depending on their existing licensing agreements, organizations can potentially reduce costs by using a bring your own license (BYOL) option, subject to Microsoft's licensing terms and conditions.

  • Replatforming to Amazon RDS for SQL Server offers the benefits of a managed service without requiring changes to existing schema or application code. This option reduces operational overhead while maintaining SQL Server compatibility. For specific use cases requiring more control, Amazon RDS Custom for SQL Server allows a hybrid approach between fully-managed and self-managed services. Additionally, the Bring Your Own Media (BYOM) feature in RDS Custom can provide potential license cost savings for eligible customers.

  • Modernizing to Amazon Aurora (or another purpose-built database service) typically yields the greatest long-term benefits. This approach frees organizations from SQL Server licensing constraints and allows them to fully use cloud-based, managed database services. While it may require more upfront effort for schema conversion and potential application changes, it often results in improved performance, scalability, and cost-effectiveness over time. The choice of target database (for example, Aurora PostgreSQL with Babelfish for SQL Server compatibility, or a completely different database type) should align with the organization's specific requirements and future technology strategy.