Q: What is AWS Outposts?

A: AWS Outposts brings AWS infrastructure, services, and operating models to virtually any data center or on-premises facility. AWS Outposts builds on the AWS Nitro system technologies that enable customers to launch EC2 instances and EBS volumes on the same AWS designed infrastructure that we use in AWS data centers. Customers can use AWS Outposts to launch and manage a wide range of AWS services such as Amazon ALB for load balancing, Amazon ECS and Amazon EKS for containers, Amazon EMR for big data, and Amazon RDS for databases, with more to be added in the future. With Outposts customers use the same AWS Management Console, Software Development Kit (SDK), and Command Line Interface (CLI) tools that they use in AWS today. 

Q: How do I get started?

A: To get started, you log into the AWS Management Console to configure and order your Outpost. You can choose from a wide range of EC2 instance types and EBS storage options to order the ideal number of instances with the desired storage for your applications. We ship Outposts directly to your desired location, where our AWS certified installation team will simply connect the equipment into your local power and network. Once connected, you can view the newly available Outposts capacity in the AWS Management Console, and can launch instances to EC2 on Outposts as part of your Amazon Virtual Private Cloud (VPC). 

Q: Why would I use AWS Outposts instead of operating in an AWS Region?

A:  AWS Outposts brings AWS infrastructure to your premises to enable latency sensitive use cases that require near real-time response and allow you to keep some data on premises for processing to avoid transmitting large volumes of data over the bandwidth to a Region. For example, AWS Outposts is ideal for telecom virtual network functions (VNFs), for low latency data services, high frequency trading platforms, high speed industrial automation, processing and editing live video streams, and traditional enterprise applications that depend on local databases.

Q: Can I use Outposts in a disconnected environment?

An Outpost continually communicates back to a parent AWS Region and must remain connected to operate. Outposts are not designed for disconnected operations. We recommend that customers have highly available networking connections back to their AWS Region.

Q: Can Outposts be used to meet data sovereignty requirements?

AWS Outposts is not designed to address specific data sovereignty requirements, or instances where you cannot send traffic off premises. 

Q: Which AWS Regions support AWS Outposts?

A: We plan to support AWS Outposts in all Regions and will have more details on availability over the course of 2019. You choose which Region and Availability Zone your Outpost is mapped to, which determines where any data sent back from the control plane or data plane will reside.

Q: Can I reuse my existing servers in an Outpost?

A: No, AWS Outposts leverages AWS designed infrastructure, and is only supported on proprietary AWS hardware that is optimized for secure, high-performance, and reliable operations.

Q: How do I use VMware Cloud on AWS on an Outpost?

A: You can create a VMware Software-Defined Data Center (SDDC) on your Outpost as part of your fully managed VMware services and environments, the same way you do on AWS today.

Q: Can I order my own hardware that can be installed as part of my Outpost rack?

A: No, AWS Outposts provides fully integrated AWS designed racks with built in top-of-rack switches and redundant power supply to ensure an ideal AWS experience. You can order as much compute and storage infrastructure as you need by selecting from the range of available Outpost options, or work with us to create a custom combination with your desired EC2 and EBS capacity. These Outpost racks are pre-validated and tested to ensure that you can get started quickly with no additional effort or configuration required on-site. We also offer a range of plug-and-play rack-level power and network configuration options to support virtually any on-premises site.

AWS Services and Networking on Outposts

Q: Which EC2 instances and EBS volume types are supported on Outposts?

A: Customers can choose from a wide selection of general purpose, compute optimized, memory optimized, and graphics optimized EC2 instance families (including M5, C5, R5, I3en, and G4) with local instance storage options, as well as EBS gp2 volumes for persistent local block storage. You can have multiple instance types within the same Outpost rack. You can start with a quarter rack of capacity and scale to half or full Outposts racks as you scale your applications.

Q: Where are EBS snapshots stored?

A: EBS snapshots associated with your Outpost EBS volumes are stored using Amazon S3 in the Region associated with your Outpost. They are sent from your Outpost over the connection back to the Availability Zone for storage.

Q: Which AWS Services are supported by AWS Outposts?

A: Customers can provision Amazon EC2 instances and EBS volumes locally on an Outpost. At launch or in the months after, we plan to add services such as Amazon RDS, Amazon ECS, Amazon EKS, AWS ALB, and Amazon EMR on Outposts. Like today, customers can use Private Link gateway endpoints to privately connect their VPC resources to regional AWS services such as Amazon S3 and DynamoDB.

Q: How do I manage deploying the same application across many Outposts?

A: You will be able to use AWS deployment tools to manage your applications across many Outposts, for example using CloudFormation templates to replicate the same architecture for many Outposts.

Q: Is there a software-only version of AWS Outposts?

A: No, AWS Outposts is a fully managed service that provides you with access to AWS capacity.

Q: Do you support container services on Outposts?

A: Yes, container services like Amazon ECS and Amazon EKS will be supported on Outposts.

Q: Do you support database services on Outposts?

A: Yes, database services like Amazon RDS will be supported on Outposts.

Q: Do you support load balancers on Outposts?

A: Yes, load balancers like AWS ALB will be supported on Outposts.

Q: How does Outposts capacity appear in my VPC?

A: You can associate a subnet to your Outpost and seamlessly extend your regional VPC environment to your on-premises facility. Outpost instances and related services will appear as part of your regional VPC, similar to an Availability Zone with associated subnets.

Q: Can I have two different workloads that shouldn’t be able to access each other on two different subnets on the same Outpost?

A: Yes - just as you would be able to have workloads on two different isolated subnets in the AWS Cloud, workloads on different subnets on the same Outpost within the same VPC will not interact.

Q: How does AWS Outposts provide low latency access to compute and storage infrastructure (without having to bounce off the Region)?

A: To enable low latency access to an Outpost from your local network, Outposts will give you the ability to map to your on-premises network with both dynamic routing (BGP) and static routing (1:1 NAT).

Q: Can I allow other accounts in my organization to launch instances on my Outpost?

A: Yes, you can share subnets and instances associated with an Outpost with one or more accounts in your organization.

Getting Started with Ordering & Installation

Q: How do I order my Outpost?

A: You will be able to order Outposts through your AWS console or API, and will have the flexibility to choose site location and delivery window. Choose from a range of pre-validated Outposts SKUs, to best align with your intended use case and compatible with the power and physical attributes of your site such as weight, uplink bandwidth, and supported site interfaces. Certified AWS Outposts personnel will deliver the Outposts to your site, connect it to power and network, and ensure Outposts capacity is available from the AWS console.

Q: Are there any prerequisites for deploying an Outpost at my location?

A: You will need to meet basic power, networking and space requirements. Outposts need 5-15 kVA, at least 1G networking uplink, and space for a 42U rack (80” X 24” X 48” dimensions). We can also work with you to identify additional form factors and configuration options that suit your sites. Given that Outposts control planes sit in the AWS Cloud, you should plan for consistent network connectivity through either a Virtual Private Network (VPN) connection or AWS Direct Connect connection.

Q: How does AWS install an Outpost rack at my facility?

A: After you place your order, you will be able to schedule a delivery date for AWS to arrive at your site and install your Outpost. You will provide basic information about your site, including power and networking availability, and specifiy any on-site personnel safety or security requirements. On your installation date, our team will arrive with the Outpost rack and move it to your identified position, then simply connect it to power and networking.

Support & Maintenance

Q: How does AWS maintain AWS Outposts infrastructure?

A:  When your Outpost comes online, AWS will monitor it as part of the public Region and will automatically execute software upgrades and patches as part of its on-going regional operations.

If there is a need to perform physical maintenance, our team will reach out to schedule a time to come to your site. Our team may replace a given unit module as appropriate - no server or switch servicing is performed on premises.

Q: What happens if an EC2 instance on an Outpost needs to be retired?

A: You will receive a maintenance notification through the EC2 and AWS Outposts console dashboards. If you have additional capacity available on your Outpost, you can migrate existing workloads to that excess capacity. You will be able to schedule instance restart using existing channels. If capacity replacements are necessary, you will be able to schedule delivery windows through the maintenance interface. You will have access to the same workload migration tools that you have access to in the AWS Cloud.

Q: What happens when my facility's network connection goes down?

A: As the AWS control plane remains in the parent Region, drops in network connectivity may impact the availability of APIs that cause mutating changes to the resources operating on the Outpost. For example, launching new instances to the Outpost and starting or stopping instances will not work without this network connectivity. However, instances and volumes on the Outpost will continue to operate normally and can be accessed locally via the local gateway.

Q: What type of control plane information flows back to the parent Region?

A: As an example, information about instance health, instance activity (launched, stopped), and the underlying hypervisor system may be sent back to the AWS parent Region. This enables the AWS Outposts service team to provide alerting on instance health and capacity, and to push out patches and updates to the Outpost. As a result, your team does not need to implement your own tooling to manage these elements, or to actively push security updates and patches. You are not able to directly monitor this traffic. During periods of disconnect, this information cannot be sent back to the parent Region.

Additional information

Q:  How can I learn more about AWS Outposts?

A: You can sign up to learn more about AWS Outposts.