Although the great outdoors may be enticing right now in the middle of Summer, there are some great announcements from AWS this month that are vying for our attention. From adding Amazon RDS capabilities to AWS Outposts and a general release of AWS IoT SiteWise, to Managed Scaling in Amazon EMR, a .NET Porting Assistant and new Graviton2 instance types, there’s lots to look at and much of it is important to the enterprise. As usual, what follows isn’t an exhaustive list of all the AWS announcements made this month, but is instead a hand-picked selection most likely to assist enterprise thought leaders working to drive cloud adoption and efficiency within their organizations.
Amazon RDS on AWS Outposts
Available now for just over half a year, AWS Outposts are physical racks that allow organizations to run cloud workloads on AWS hardware in their own data centers. This lets customers accommodate workloads that have ultra low latency requirements or data residency concerns. Until now this has been primarily limited to running Amazon EC2-based workloads on AWS Outposts, however as of this month the Amazon Relational Database Service (RDS) is now available, allowing organizations to take advantage of AWS’ managed service for both MySQL and PostgreSQL based databases, with more engines to come in the future.
Much like AWS Outposts, Amazon RDS on AWS Outposts is ideal for cases where organizations require local access to the databases, particularly in manufacturing and IoT applications, or to ensure data residency for compliance or security reasons.
While the general usage of Amazon RDS on AWS Outposts is the same as if it were directly in the AWS Cloud, there are some nuances to keep in mind, mainly that read replica and HA cluster functionality is not available for Amazon RDS on AWS Outposts, and that engine versions currently available for MySQL and PostgreSQL are limited. That said, this provides a powerful new functionality on AWS Outposts, simplifying the creation and management of databases massively.
For more details, check out the AWS blog post or service page.
AWS IoT SiteWise General Availability
AWS IoT SiteWise has spent the last couple of years in preview as AWS honed its capabilities and features; this month it hits prime time, letting organizations easily collect and centralize the information and metrics from their industrial equipment. By using AWS IoT Sitewise to organize and manage the data, organizations can use it to power monitoring and business decisions across their equipment installations using queries and alerts.
AWS IoT SiteWise uses a gateway running in the organization’s on-premises environment to connect to the data stores used by the equipment and securely collect it into AWS, where it can then be streamed in near real-time to power dashboards or applications.
The ability to collect data from multiple locations and equipment types, and to centralize it in a consistent fashion means that organizations can better understand their global operations with greater ease, and make well-informed decisions quickly to react to their needs and challenges.
More information on AWS IoT SiteWise can be found here, as well as a walkthrough of some of its features in the AWS blog post.
Managed Scaling for Amazon EMR
Amazon Elastic Map Reduce (EMR) is a valuable service allowing organizations to process large amounts of data leveraging open source tools like Apache Spark, Apache HBase, and Presto. Typically in the past, the scale of the Amazon EMR cluster had to be defined ahead of time, or using complex and custom written scaling rules, resulting in difficult to manage clusters or cost inefficiencies due to overprovisioning. Now with Amazon EMR Managed Scaling, AWS uses constant sampling of key metrics to understand how the workloads are using compute and whether the cluster needs to scale up or down. By providing a minimum and maximum value for compute the service automatically scales within those values. Work is assigned to the cluster or is completed, helping to avoid cost overruns while ensuring the optimal use of the deployed resources.
The Amazon EMR Managed Scaling algorithm is able to monitor the cluster at a much higher frequency than regularly configured autoscaling, making for much more precise scaling actions as well as requiring less overhead and management from your organization. Finally, Amazon EMR Managed Scaling can be used with Instance Fleets to enable the use of Spot Instances in conjunction with On-Demand, allowing for even greater cost savings.
Check out the AWS blog for examples and more details.
.NET Porting Assistant
With .NET Framework 4.8 being the last major version of the .NET Framework many organizations are looking to migrate their .NET Framework applications to the more modern .NET Core, but are hampered by it typically being an extremely laborious process as developers identify and update dependencies, and replace APIs that are no longer available in .NET Core. To alleviate this process, AWS has released the .NET Porting Assistant, which can analyze both the application code as well as the full dependency tree. It provides recommendations and guides to replacements, as well as learns over time to improve its recommendations based on usage patterns and missing package/API occurrences across AWS’s broad customer base.
By utilizing the .NET Porting Assistant, organizations can quickly move to modernize their .NET stack, as well as to be able to take advantage of performance gains in .NET core, leverage containerization capabilities for easier scaling and resource utilization, and, importantly, reduce licensing costs by moving to Linux based instances.
The AWS blog has a great sample walk through, and more information on the service page itself.
AWS Graviton2 Amazon EC2 Instances with NVMe Local Storage
Finally, AWS continues to expand their AWS Graviton2 Amazon EC2 Instances, which use Arm-based processors and the AWS Nitro System custom hardware and software, to deliver better price performance than their x86-based cousins.
This month’s announcement of the “d” variant (M6gd, C6gd, and R6gd) provides fast, locally attached SSD drives which are ideal for temporary file processing or caching – as with SSD elsewhere, the data only persists for the life of the instance. AWS’s storage offering for these instance types are up to 2 x 1.9TB, depending on the instance sizes. The Graviton2 “d” variant offers 50% more storage GB/vCPU than their non-Graviton2 based equivalents, the M5d, C5d, and R5d. Many organizations are already leveraging the Graviton2 instance types to see increased performance, allowing them to reduce fleet size and thereby reduce cost.
More information can be found in the AWS blog post, and AWS has created a Git repository to provide help on designing code for Arm-based processors.
To follow these monthly updates and gain insights on how they can impact your business, subscribe to our blog.