Services supported locally on the new AWS Outposts will include Amazon ECS and Amazon EKS clusters for container-based applications, Amazon EMR clusters for data analytics and Amazon RDS instances for relational database services when Outposts launches later this year.
Amazon SageMaker and Amazon MSK are among other services that will follow soon after, according to Matt Garman, vice president of compute services at Amazon Web Services.
The fully managed AWS Outposts, which essentially extends an AWS region into customers’ own data centres, was announced at the AWS re:Invent conference last November. It will include two versions, VMware Cloud on AWS Outposts and the native AWS Outposts version.
AWS Outposts allows customers to use native AWS services, infrastructure, application programming interfaces (APIs) and other tools in their on-premises facilities, with integrated data center hardware that runs in the VMware or native AWS environments. The new fully managed service is designed for running applications with low latency and local data-processing requirements.
Customers can use AWS Outposts to launch a range of Amazon EC2 instances — C5, M5, R5, I3en and G4, with or without local storage options — and Amazon Elastic Block Store volumes locally, according to Garman.
“You can use private connectivity to your Amazon S3 buckets or Amazon DynamoDB tables in the public region,” Garman wrote in a blog post yesterday that offers more clarity on Outposts. “Amazon tools will work with Outposts as well. API calls will be logged via CloudTrail automatically, and existing CloudFormation templates will work. When AWS launches new innovations, they will work with Outposts so customers can always take advantage of the latest technologies.”
AWS declined to name customers or partners that have early access to Outposts, but Garman cited “immense” customer interest as part of their hybrid cloud strategies. AWS is speaking with companies in industries including healthcare, financial services, manufacturing, media and entertainment, and telecommunications, he said.
“One of the most common scenarios is applications that need single-digit millisecond latency to end-users or onsite equipment,” Garman said. “Customers may need to run compute-intensive workloads on their manufacturing factory floors with precision and quality. Others have graphics-intensive applications such as image analysis that need low-latency access to end-users or storage-intensive workloads that collect and process hundreds of (terabytes) of data a day.”
One unnamed early Outposts user is using it to control and operate industrial equipment at hundreds of work sites around the world.
“They already run centralised decision-making applications in AWS to identify what work to execute at which site,” Garman said. “Predictable low-latency access to local compute resources is essential for their on-premises control systems to manage materials with smoothness and speed. For instance, control systems need to process video streams to sense the product on the conveyor belt and execute a robotic movement to direct the product to the right location. Their sites also run video monitoring applications where the captured data can exceed available bandwidth (when) they want to conduct video encoding on-premises.”
After connecting that customer’s Outpost to the nearest local AWS region, the company has complete control over its virtual network, including selection of an internet protocol (IP) address range, the creation of subnets and the configuration of route tables and network gateways, according to Garman.
“Using Outposts, the customer plans to standardize tooling across on-premises and the cloud, and automate deployments and configurations across hundreds of sites by using the same APIs, the same IAM (identity and access management) permissions, the same EC2 AMIs (Amazon machine images), the same CloudFormation templates and the same deployment pipelines everywhere,” Garman said.
AWS Outpost racks
The Outposts compute and storage data center racks will be built with AWS-designed hardware. AWS will deliver, install, operate and monitor the physical Outposts infrastructure, and provide automatic updates and patches.
The racks, which come fully assembled, include the same hardware that AWS uses within its public region data centers. They’re 24 inches wide, 48 inches deep and 80 inches tall, and come with casters for maneuverability. They have a bus bar in the back and a power shelf in the middle, and use a centralised redundant power- conversion unit and a DC distribution system in the back plane.
Every active component is redundant and can be removed and replaced without impacting other workloads, according to AWS.
Outpost racks use the same control plane that EC2 racks use, but they include an additional Nitro chip in every server to help connect them back to the public AWS region.
While the hardware racks will include built-in, top-of-rack switches, AWS partners don’t see that as an AWS move to take on network switch market leader Cisco Systems or other manufacturers.
AWS isn’t disclosing the individual components used for AWS Outposts racks, but all have been thoroughly tested and verified by AWS, according to a source close to the company.
“Customers are less concerned about the individual hardware components and more interested in the value that the consistent AWS experience and the fully managed service that AWS Outposts delivers,” the source said.