AWS Documentation » AWS Snowball » Developer Guide » Using an AWS Snowball Edge » Using Amazon EC2 Compute Instances

This guide is for the Snowball Edge. If you are looking for documentation for the Snowball, see the AWS Snowball User Guide.

Using Amazon EC2 Compute Instances

The following contains an overview on using compute instances on a Snowball Edge device, including conceptual information, procedures, and examples.

Overview

Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides resizeable computing capacity—literally, servers in Amazon's data centers—that you use to build and host your software systems. You can now run Amazon EC2 compute instances hosted on a Snowball Edge with the new sbe1.xxxx instance type. This provides more options to run your compute workloads, adding to the existing support for serverless compute through AWS Lambda powered by AWS Greengrass.

The compute instance types supported for use on a Snowball Edge (identified by sbe1.xxxx) are unique to Snowball Edge devices. Like their cloud-based counterparts, these instances require Amazon Machine Images (AMIs) to launch. You choose the AMI that will be that base image for an sbe1.xxxx instance in the cloud, before you create your Snowball Edge job.

To use a compute instance on a Snowball Edge, create a job and specify your AMIs. This can be done from the AWS Snowball Management Console, with the AWS CLI, or with one of the AWS SDKs. Typically, there are some housekeeping prerequisites that you must resolve before creating your job, in order to use your instances.

After you've handled the prerequisites, created your job, and waited for your device to arrive, you can start managing your AMIs and instances. You can manage your compute instances on a Snowball Edge through an Amazon EC2–compatible endpoint. This type of endpoint supports many of the Amazon EC2 AWS CLI commands and actions for the AWS SDKs. Note that you can't use the AWS Management Console on the Snowball Edge to manage your AMIs and compute instances.

When you're done with your device, return it to AWS. If the device was used in an import job, the data transferred using the Amazon S3 Adapter for Snowball or the file interface will be imported into Amazon S3. Otherwise, we perform a complete erasure of the device when it is returned to AWS. This erasure follows the National Institute of Standards and Technology (NIST) 800-88 standards.

Important

Data in compute instances running on a Snowball Edge will not be imported into AWS.

Compute Instances on Clusters

You can use compute instances on clusters of Snowball Edge devices, and the procedures and guidance for doing so is the same as it would be for using compute instances on a stand-alone device.

When you create a cluster job with AMIs, a copy of each AMI will exist on each node in the cluster. For that reason, there can only be 10 AMIs associated with a cluster of devices, regardless of the number of nodes on the cluster. When you launch an instance in a cluster, you declare the node that will host the instance in your command, and the instance will run on a single node.

Pricing for compute instances on Snowball Edge

There are additional costs associated with using compute instances. For more information, see AWS Snowball Edge Pricing.

Prerequisites

Before creating your job, keep the following information in mind.

• Before you add any AMIs to your job creation request, you'll need to have an AMI created in your AWS account of the supported image type. Currently, supported AMIs are based on the CentOS 7 (x86_64) - with Updates HVM, Ubuntu Server 14.04 LTS (HVM), and Ubuntu 16.04 LTS - Xenial (HVM) images. You can get any one of these images from the AWS Marketplace.

• All AMIs must be Amazon EBS based, with a single volume.

• If you are planning connecting to a compute instance running on a Snowball Edge, you must use SecureShell (SSH). To do so, you'll first need to bake-in the key pair. For more information, see (Optional) Configure an AMI so you can SSH into the Compute Instances That the AMIs Launch on the Device.

Limits for Compute Instances on a Snowball Edge

Supported AWS Regions

You can add compute instances to your devices in the following regions:

• US East (Ohio)

• US East (N. Virginia)

• US West (N. California)

• US West (Oregon)

• Asia Pacific (Singapore)

• Asia Pacific (Sydney)

• Asia Pacific (Tokyo)

• EU (Frankfurt)

• EU (Ireland)

• EU (London)

• EU (Paris)

• South America (São Paulo)

Storage Limits

The storage available for compute resources is a separate resource from the dedicated Amazon S3 storage on a Snowball Edge device. The total available storage for sbe1.xxxx instances and AMIs is 1000 GiB. The total available storage for Amazon S3 on a standalone device is based on the following factors.

• AWS Snowball Edge device without compute instances – 80 TB

• AWS Snowball Edge device with compute instances – 70 - 60 TB

When you create a job with compute instances, 10 TB of space are reserved for your AMIs. This 10 TB is allocated regardless of the number and size of AMIs associated with your job. When you start the device for the first time, the actual storage required for each of your AMIs is also reserved. So if you have 5 TB of AMIs, when you turn on your device for the first time, it will have 65 TB of available space (10 TB allocated for compute instances, and 5 TB reserved for your actual AMIs).

There's also a limit to the physical resources available for compute instances on a device. To find the resource requirements and limitations for this feature, see the following tables.

Feature	Limitation
Number of AMIs on a single Snowball Edge	10
Number of volumes per instance	1
Concurrently running (or stopped) instances	Varies depending on available resources

Instance type	vCPU cores	Memory (GiB)
sbe1.small	1	1
sbe1.medium	1	2
sbe1.large	2	4
sbe1.xlarge	4	8
sbe1.2xlarge	8	16
sbe1.4xlarge	16	32

Shared Compute Resource Limitations

The vCPU cores, memory, and storage available for compute resources is a separate resource from the dedicated Amazon S3 storage on a Snowball Edge. However, compute resources are spread across the following services on the device:

• File interface – while ACTIVE, this service uses 8 vCPU cores and 16 GiB of memory.

• AWS Greengrass and AWS Lambda powered by AWS Greengrass – while ACTIVE, these services combined use 4 vCPU cores and 8 GiB of memory.

• Compute instances – If no other compute services are ACTIVE, you have up to 16 vCPUs and 32 GiB of memory for your compute instances.

You can determine whether or not a service is ACTIVE on a Snowball Edge using the Snowball client command snowballEdge describe-service. For more information, see Getting Service Status.

Important

Snowball Edge devices with any of it's available compute resources maxed out will be unable to launch new compute resources, and those requests will fail. For example, if you try to start the file interface while also running a sbe1.4xlarge compute instance, the file interface will not start.

Creating a Job with Compute Instances

In this section, you create your first compute instance job. This process has some optional steps in the beginning, depending on whether or not you plan on using SSH to connect to your instances when they're running on your Snowball Edge device.

Important

Be aware of the following points before you create your job:

• Make sure the vCPU, memory, and storage values associated with your AMI match the type of instance you want to create.

• If you're going to use SSH to connect to the instance after you launch the instance on your Snowball Edge, you must first follow this optional procedure. You can't update the AMIs on your Snowball Edge after the fact. You must do this step before creating the job.

(Optional) Configure an AMI so you can SSH into the Compute Instances That the AMIs Launch on the Device

In order to use SSH to connect to your compute instances on Snowball Edge devices, you must perform the following optional procedure to bake-in the SSH key into the AMI before creating your job. We also recommend that you use this procedure to set up your applications on the instance that you'll use as the AMI for your job.

Important

If you don't follow this optional procedure, there is no way to SSH into your instances when you receive your Snowball Edge device.

1. Launch a new instance in the AWS cloud based on the CentOS 7 (x86_64) - with Updates HVM, Ubuntu Server 14.04 LTS (HVM), or Ubuntu 16.04 LTS - Xenial (HVM) images.

   When you launch your instance, you'll want to make sure that the storage size that you assign to the instance is appropriate for your later use on the Snowball Edge. In the Amazon EC2 AWS Management Console, this is done in Step 4: Add Storage. For a list of the supported sizes for compute instance storage volumes on a Snowball Edge, see Limits for Compute Instances on a Snowball Edge.

2. Install and configure the applications that you want to run on the Snowball Edge, and test that they work as expected.

3. Make a copy of the PEM/PPK file that you used for the SSH key pair to create this instance. You'll need to save this file to the server that you'll use to communicate with the Snowball Edge. This file is required for using SSH to connect to the launched instance on your device, so make a note of the path to this file.

4. Save the instance as an AMI. For more information, see Creating an Amazon EBS-Backed Linux AMI in the Amazon EC2 User Guide for Linux Instances.

5. Repeat this procedure for each of the instances that you want to SSH into, keeping copies of your different SSH key pairs and taking note of the AMIs they're associated with.

Creating Your Job in the Console

Your next step will be to create a job. Your job can be of any job type, including a cluster. Using the AWS Snowball Management Console, and follow the instructions provided in Create Your First Job, add these additional steps when you get to the Step 3: Give job details page in the job creation wizard.

1. Choose Enable compute with EC2.

2. Choose Add an AMI.

3. In the dialog box that opens, choose an AMI and choose Save.

4. Add up to 10 total AMIs to your job.

5. Continue creating your job as normal.

Creating Your Job in the AWS CLI

Alternatively, you can create your job using the AWS CLI. To do this, open a terminal and run the following command, replacing the red text with your actual values:

aws snowball create-job --job-type IMPORT --resources '{"S3Resources":[{"BucketArn":"arn:aws:s3:::bucket-name"}],"Ec2AmiResources":[{"AmiId":"ami-12345678"}]}' --description Example --address-id ADIEXAMPLE60-1234-1234-5678-41fEXAMPLE57  --kms-key-arn arn:aws:kms:us-west-2:012345678901:key/eEXAMPLE-1234-1234-5678-5b4EXAMPLE8e --role-arn arn:aws:iam::012345678901:role/snowball-local-s3-lambda-us-west-2-role --snowball-capacity-preference T100 --shipping-option SECOND_DAY --snowball-type EDGE

After it arrives and you've unlocked your device, you'll need to use the Snowball client to get your local credentials. For more information, see Getting Credentials.

Network Configuration for Compute Instances

After you’ve launched your compute instances on a Snowball Edge device, you need to set up your network configuration for each instance. Use the following procedure whenever you launch a new instance to give that instance a virtual network interface and IP address.

1. Make sure there's power to your device and that one of your physical network interfaces, like the RJ45 port, are connected with an IP address.

2. Get the IP address associated with the physical network interface that you're using on the Snowball Edge. For more information, see Connect to Your Local Network.

3. Configure your Snowball client. For more information, see Configuring a Profile for the Snowball Client.

4. Unlock the device. For more information, see Unlocking AWS Snowball Edge Devices.

5. Run the snowballEdge describe-device command to get the list of physical network interface IDs.

6. Identify the ID for the physical network interface that you want to use, and make a note of it.

7. Run the snowballEdge create-virtual-network-interface command. The following examples show running this command with the two different IP address assignment methods, either DHCP or STATIC.

   snowballEdge create-virtual-network-interface \
   --physical-network-interface-id s.ni-abcd1234 \
   --ip-address-assignment DHCP
            
           //OR//
           
   snowballEdge create-virtual-network-interface \
   --physical-network-interface-id s.ni-abcd1234 \
   --ip-address-assignment STATIC \
   --static-ip-address-configuration IpAddress=192.0.2.0,Netmask=255.255.255.0

8. The command returns a JSON structure that includes the IP address. Make a note of that IP address as you'll need it for the ec2 associate-address AWS CLI command.

9. Anytime you need this IP address, you can use the snowballEdge describe-virtual-network-interfaces Snowball client command, or the aws ec2 describe-addresses AWS CLI command to get it.

10. Associate your newly created IP address with the instance, using the following AWS CLI command:

    aws ec2 associate-address --public-ip 192.0.2.0 --instance-id s.i-01234567890123456 --endpoint <physical IP address for your Snowball Edge>:8008

Connecting to Your Compute Instance on a Snowball Edge using SSH

In order to use SSH to connect to your compute instances on Snowball Edge devices, you must have followed the optional procedure to bake-in the SSH key into the AMI before creating your job. For more information on that procedure, see (Optional) Configure an AMI so you can SSH into the Compute Instances That the AMIs Launch on the Device. If you haven't followed that procedure, there is no way to SSH into your instances.

To connect to your instance with SSH

1. Make sure your device is powered on, connected to network, and unlocked. For more information, see Connect to Your Local Network.

2. Make sure you have your network settings configured for your compute instances. For more information, see Network Configuration for Compute Instances.

3. Refer to your notes to find the PEM/PPK key pair that you used for this specific instance, and make a copy of those files somewhere on your computer. Make a note of the path to the PEM file.

4. Connect to your instance through SSH as in the following example command. Note that the IP address is the IP address of the VNIC that you set up in Network Configuration for Compute Instances.

   ssh -i path/to/PEM/key/file instance-user-name@192.0.2.0

   For more information, see Connecting to Your Linux Instance Using SSH in the Amazon EC2 User Guide for Linux Instances.

Transferring Data from Amazon EC2 Compute Instances to Amazon S3 Buckets on the Same Snowball Edge

You can transfer data between compute instances and Amazon S3 buckets on the same Snowball Edge device. This is done by using the supported AWS CLI commands and the appropriate endpoints. For example, assume that you want to move data from a directory in my sbe1.xlarge instance into the Amazon S3 bucket, myBucket on the same device, using the Amazon S3 endpoint 192.0.2.1:8080. You would use the following procedure.

Note

This procedure only works if you've followed the instructions in (Optional) Configure an AMI so you can SSH into the Compute Instances That the AMIs Launch on the Device.

To transfer data between a compute instance and a bucket on the same Snowball Edge

1. Use SSH to connect to your compute instance.

2. Download and install the AWS CLI. If your instance doesn't already have the AWS CLI, you'll need to download and install it. For more information, see Installing the AWS Command Line Interface.

3. Configure the AWS CLI on your compute instance to work with the Amazon S3 endpoint on the Snowball Edge. For more information, see Getting and Using Local Amazon S3 Credentials.

4. Use the supported Amazon S3 AWS CLI commands to transfer data. For example:

   aws s3 cp ~/june2018/results s3://myBucket/june2018/results --recursive --endpoint http://192.0.2.1:8080