AWS Database Blog

New, Memory-Optimized Amazon EC2 Instance Types Drive Database Workloads

Perfectly sized instances for maximizing Microsoft SQL Server Standard

In September, Amazon Web Services announced availability of the new Amazon EC2 x1e.32xlarge instance type with 128 vCPU and 3,904 GiB of memory.

Since that announcement, we have heard from customers that they want more instance configuration choices with fewer vCPUs, while maintaining a high ratio to memory. Last week, the introduction of the x1e family was completed with the announcement of general availability of five new sizes—x1e.xlarge, x1e.2xlarge, x1e.4xlarge, x1e.8xlarge, and x1e.16xlarge. These are currently available in the US East (N. Virginia), US West (Oregon), EU (Ireland), Asia Pacific (Tokyo), and Asia Pacific (Sydney) regions.

With the highest memory per vCPU among Amazon EC2 instance types and one of the lowest prices per GiB of memory, the new x1e instance sizes are well suited for high-performance databases, in-memory databases, and other memory-optimized enterprise applications.

Bill Ramos, director of technical product management at DB Best, has this to say about the new x1e instance types:

The new x1e memory-optimized instance family hits a sweet spot with our SQL Server 2016/2017 customers who want in-memory computing using Standard Edition. With the new SQL Server 2017 read-scale availability groups, our customers can perform data warehouse like queries using in-memory clustered columnstore performance while running the online transaction processing on the primary replica using Standard Edition as well. It’s great that you can start with a 4 core, 122 GiB system using SQL Server 2017 Standard Edition and then scale up as needed. With the 8 core, 244 GiB system, customers can run their SQL Server database instance with 128 GiB with Analysis Services using another 64 Gib and still have room for other applications all using Standard Edition.

The smallest member of the x1e family (x1e.xlarge) has 4 vCPU and 122 GiB memory, and the x1e.2xlarge has 8 vCPU and 244 GiB memory. These are ideal candidates for Microsoft SQL Server Standard Edition. With the x1e.2xlarge instance type, you can allocate the maximum allowed memory (128 GB) for SQL Server Standard database engine and still have enough remaining for Analysis, Integration, or Reporting Services (SSAS, SSIS, or SSRS).

These new instance types, along with the SIOS DataKeeper Quick Start introduced in May and the price reduction announced in July for SQL Server Standard on Amazon EC2, bring more affordable SQL Server high availability to AWS customers. With the lowest price per GiB of memory among Amazon EC2 instance types, you can take advantage of not only large amount of memory, but also optimize your spend. For example, the on-demand price for an x1e.xlarge instance running Microsoft Windows and SQL Server Standard Edition is just 61 percent more than the cost of an r4.xlarge instance but with four times more memory. For more pricing comparisons, take a look at the AWS Simple Monthly Calculator.

With x1e instances, you still get all the licensing options for running SQL Server on Amazon EC2. With the license included SQL Server option, you can avoid making a long-term purchase and don’t have to deal with true-ups, software compliance audits, or Software Assurance. This option works well if you prefer to avoid buying licenses and want to upgrade from an older version of SQL Server.

You can also choose the License Mobility option. With this option, you can use your active Software Assurance agreement to bring your existing licenses to EC2 without needing dedicated infrastructure.

You can choose to bring you own licenses to the new X1e instances and take advantage of your existing license investment while further optimizing your upgrade costs. You can run SQL Server on EC2 Dedicated Instances or EC2 Dedicated Hosts, with the potential to reduce operating costs by licensing SQL Server on a per-core basis. You can bring your own SQL Server 2017 licenses (BYOL) to EC2 Windows, EC2 Linux instances, or to Docker containers running in Amazon EC2.

About the Author

Tom Staab is a partner solutions architect at Amazon Web Services. He works with our customers to provide guidance and technical assistance on database projects, helping them improving the value of their solutions when using AWS.