Replication server settings reference - Application Migration Service

Replication server settings reference

Replication servers are lightweight Amazon EC2 instances that are used to replicate data between your source servers and AWS. Replication servers are automatically launched and terminated as needed. You can modify the behavior of the replication servers by modifying the settings for a single source server or multiple source servers. Alternatively, you can run AWS Application Migration Service with the default replication server settings.

You can configure a variety of replication server options, including:

  • The subnet within which the replication server will be launched

  • Replication Server instance type

  • Amazon EBS volume types

  • Amazon EBS encryption

  • Security groups

Once you have finalized your changes, click the orange Save replication settings button on the bottom right of the page.

Replication settings interface showing server, data routing, and resource configuration options.

Staging area subnet

Choose the Staging area subnet that you want to allocate as the staging area subnet for all of your replication servers.

Dropdown menu for selecting a staging area subnet in the replication servers section.

The best practice is to create a single dedicated, separate subnet for all of your migration waves using your AWS account. Learn more about creating subnets in this AWS VPC article.

If a default subnet does not exist, select a specific subnet. The drop-down menu contains a list of all subnets that are available in the current AWS Region.

Dropdown menu showing subnet options including Private and Public subnets in a VPC.
Note

Changing the subnet does not significantly interfere with ongoing data replication, although there may be a minor delay of several minutes while the servers are moved from one subnet to another.

Using multiple subnets

The best practice is to use a single staging area subnet for all of your migration waves within a single AWS account. You may want to use multiple subnets in certain cases, such as the migration of thousands of servers.

Note

Using more than one staging area subnet might result in higher compute consumption as more replication servers will be needed.

Launching replication servers in Availability Zones

If you want your replication servers to be launched in a specific Availability Zone, then select or create a subnet in that specific Availability Zone. Learn more about using Availability Zones in this Amazon Elastic Compute Cloud article.

Replication server instance type

Choose the Replication server instance type. This will determine the instance type and size that will be used for the launch of each replication server.

Dropdown menu for selecting replication server instance type, with m5.xlarge option shown.

The best practice is to not change the default replication server instance type unless there is a business need for doing so.

By default, AWS Application Migration Service utilizes the t3.small instance type. This is the most cost effective instance type and should work well for most common workloads. You can change the replication server instance type to speed up the initial sync of data from your source servers to AWS. Changing the instance type will likely lead to increased compute costs.

You can change the Replication server instance type according to your preferences. The drop-down menu contains all available instance types. Recommended and commonly used instance types are displayed first.

You can search for a specific instance type within the search box.

The replication server instance type can be changed for servers that are replicating too slowly or servers that are constantly busy or experience frequent spikes. These are the most common instance type changes that are made:

  • Servers with less than 26 disks – Change the instance type to m5.large. Increase the instance type to m5.xlarge or higher as needed.

  • Servers with more than 26 disks (or servers in AWS Regions that do not support m5 instance types) – Change the instance type to m4.large. Increase to m4.xlarge or higher, as needed.

Note
  • Changing the replication server instance type will not affect data replication. Data replication will automatically continue from where it left off, using the new instance type you selected.

  • By default, replication servers are automatically assigned a public IP address from Amazon's public IP space.

  • Replication Servers are only supported on x86_64 CPU architecture instance types.

Dedicated instance for replication server

Choose whether you would like to use a Dedicated instance for replication server.

Dropdown menu for selecting a dedicated Replication Server instance option.

When an external server is very write-intensive, the replication of data from its disks to a shared Replication Server can interfere with the data replication of other servers. In these cases you should choose the Use dedicated replication server option (and also consider changing Replication server instance type).

Otherwise, choose the Do not use dedicated replication server option.

Note

Using a dedicated replication server may increase the Amazon EC2 cost you incur during replication.

Amazon EBS volume type

Choose the default Amazon Amazon EBS volume type to be used by the replication servers for large disks.

Dropdown menu showing EBS volume type selection for large disks with throughput-optimized HDD option.

Each disk has minimum and maximum sizes and varying performance metrics and pricing. Learn more about Amazon EBS volume types in this Amazon EBS article.

The best practice is to not change the default Amazon EBS volume type, unless there is a business need for doing so.

Note

This option only affects disks over 500 GiB (by default, smaller disks always use Magnetic HDD volumes).

The default Lower cost, Throughput Optimized HDD (st1) option utilizes slower, less expensive disks.

You may want to use this option if:

  • You want to keep costs low

  • Your large disks do not change frequently

  • You are not concerned with how long the initial sync process will take

The Faster, General Purpose SSD (gp3) option utilizes faster, but more expensive disks.

Dropdown menu showing "Faster, General Purpose SSD (gp2)" as the selected option.

You may want to use this option if:

  • Your source server has disks with a high write rate or if you want faster performance in general

  • You want to speed up the initial sync process

  • You are willing to pay more for speed

Note

You can customize the Amazon EBS volume type used by each disk within each source server in that source server's settings. Learn more about changing individual source server volume types.

Amazon EBS encryption

Choose whether to use the default or custom Amazon EBS encryption. This option will encrypt your replicated data at rest on the Staging Area Subnet disks and the replicated disks.

  • Default – The default Amazon EBS encryption Volume Encryption Key will be used (which can be an Amazon EBS-managed key or a CMK). .

  • Custom – You will need to enter a custom customer-managed key (CMK) in the regular key ID format.

EBS encryption options showing Default and Custom choices in a dropdown menu.

If you select the Custom option, the EBS encryption key box will appear. Enter the ARN or key ID of a customer-managed CMK from your account or another AWS account. Enter the encryption key (such as a cross-account KMS key) in the regular key ID format (KMS key example: 123abcd-12ab-34cd-56ef-1234567890ab).

To create a new AWS Key Management Service key, click Create an AWS KMS key. You will be redirected to the Key Management Service (KMS) Console where you can create a new key to use.

EBS encryption settings with dropdown for custom encryption and option to create AWS KMS key.

Learn more about Amazon EBS Volume Encryption in this Amazon EBS article.

Important

Reversing the encryption option after data replication has started will cause data replication to start from the beginning.

Using an AWS KMS Customer Managed Key (CMK) for encryption

If you decide to use a Customer Managed Key (CMK), or if your default Amazon EBS encryption key is a CMK, you will need to add additional permissions to the key to allow AWS Application Migration Service to use it.

To modify the existing key policy using the AWS Management Console policy view.

  1. Navigate to the AWS KMS Console and select the AWS KMS key you plan to use with AWS MGN.

  2. Scroll to Key policy and click Switch to policy view.

  3. Click Edit and add the following JSON statements to the Statement field.

    { "Sid": "Allow AWS Services permission to describe a customer managed key for encryption purposes", "Effect": "Allow", "Principal": { "AWS": "arn:aws:iam::$ACCOUNT_ID:root" }, "Action": [ "kms:DescribeKey" ], "Resource": "*", "Condition": { "StringEquals": { "kms:CallerAccount": [ "$ACCOUNT_ID" ] }, "Bool": { "aws:ViaAWSService": "true" } } }, { "Sid": "Allow AWS MGN permissions to use a customer managed key for EBS encryption", "Effect": "Allow", "Principal": { "AWS": "arn:aws:iam::$ACCOUNT_ID:root" }, "Action": [ "kms:CreateGrant" ], "Resource": "*", "Condition": { "StringEquals": { "kms:CallerAccount": [ "$ACCOUNT_ID" ], "kms:GranteePrincipal": [ "arn:aws:iam::$ACCOUNT_ID:role/aws-service-role/mgn.amazonaws.com/AWSServiceRoleForApplicationMigrationService" ] }, "ForAllValues:StringEquals": { "kms:GrantOperations": [ "CreateGrant", "DescribeKey", "Encrypt", "Decrypt", "GenerateDataKey", "GenerateDataKeyWithoutPlaintext" ] }, "Bool": { "aws:ViaAWSService": "true" } } }, { "Sid": "Allow EC2 to use this key on behalf of the current AWS Application Migration Service user, during target launches", "Effect": "Allow", "Principal": { "AWS": [ "arn:aws:iam::$ACCOUNT_ID:root", "arn:aws:iam::$ACCOUNT_ID:role/aws-service-role/mgn.amazonaws.com/AWSServiceRoleForApplicationMigrationService" ] }, "Action": [ "kms:ReEncrypt*", "kms:GenerateDataKey*" ], "Resource": "*", "Condition": { "StringEquals": { "kms:CallerAccount": [ "$ACCOUNT_ID" ], "kms:ViaService": "ec2.$REGION.amazonaws.com" } } }
    Important
    • Replace $ACCOUNT_ID" with the AWS account ID you are migrating into.

    • Replace $REGION with the AWS Region you are migrating into.

    • The last statement can be made stricter by ensuring the principal refers to users who are going to perform StartTest or StartCutover API calls

  4. Click Save changes.

Note

If you are using a Customer Managed Key (CMK) from another account, you need to take an additional step from within that account to allow the service to leverage the CMK.

From the account in which you want to stage MGN replication servers, create a grant that delegates the relevant permissions to the appropriate service-linked role. The Grantee Principal element of the grant is the ARN of the appropriate service-linked role. The key-id is the ARN of the key.

The following is an example create-grant CLI command that gives the service-linked role named AWSServiceRoleForApplicationMigrationService in account 111122223333 permissions to use the customer-managed key in account 444455556666.

aws kms create-grant \

--region us-west-2 \

--key-id arn:aws:kms:us-west-2:444455556666:key/1a2b3c4d-5e6f-1a2b-3c4d-5e6f1a2b3c4d \

--grantee-principal arn:aws:iam::111122223333:role/aws-service-role/mgn.amazonaws.com/AWSServiceRoleForApplicationMigrationService \

--operations "Encrypt" "Decrypt" "ReEncryptFrom" "ReEncryptTo" "GenerateDataKey" "GenerateDataKeyWithoutPlaintext" "DescribeKey" "CreateGrant"

For this command to succeed, the user making the request must have permissions for the CreateGrant action.

Always use Application Migration Service security group

Choose whether you would like to Always use the Application Migration Service security group.

A security group acts as a virtual firewall, which controls the inbound and outbound traffic of the staging area subnet.

Security group settings for Application Migration Service with additional options.

The best practice is to have AWS Application Migration Service automatically attach and monitor the default Application Migration Service Security Group. This group opens inbound TCP Port 1500 for receiving the transferred replicated data. When the default Application Migration Service Security Group is activated, Application Migration Service will constantly monitor whether the rules within this security group are enforced, in order to maintain uninterrupted data replication. If these rules are altered, Application Migration Service will automatically fix the issue.

Select the Always use Application Migration Service security group option to allow data to flow from your source servers to the replication servers, and that the replication servers can communicate their state to the AWS Application Migration Service servers.

Otherwise, select the Do not use Application Migration Service security group option. Selecting this option is not recommended.

Additional security groups can be chosen from the Additional security groups dropdown. The list of available security groups changes according to the Staging area subnet you selected.

Dropdown menu showing "default" as the only available security group option.

You can search for a specific security group within the search box.

You can add security groups via the AWS Management Console, and they will appear on the security group drop-down list in the AWS Application Migration Service Console. Learn more about AWS security groups in this VPC article.

You can use the default Application Migration Service security group, or you can select another security group. However, take into consideration that any selected security group that is not the Application Migration Service default, will be added to the default group, since the default security group is essential for the operation of Application Migration Service.

Data routing and throttling

AWS Application Migration Service allows you to control how data is routed from your source servers to the replication servers on AWS through the Data routing and throttling settings.

Data routing and throttling options for private IP, public IP, and network bandwidth.

By default, data is sent from the source servers to the replication servers over the public internet, using the public IP that was automatically assigned to the replication servers. Transferred data is always encrypted in transit.

Note

The Data routing and throttling view differs slightly between the replication template view and the individual source server replication settings view, but the instructions apply to both views.

Use private IP for data replication

Choose the Use private IP option if you want to route the replicated data from your source servers to the staging area subnet through a private network with a VPN, AWS Direct Connect, VPC peering, or another type of existing private connection.

Choose Do not use private IP if you do not want to route the replicated data through a private network.

Radio button options for using or not using private IP for data replication.
Important

Data replication will not work unless you have already set up the VPN, AWS Direct Connect, or VPC peering in the AWS Console.

Note
  • If you selected the Default subnet, it is highly unlikely that the private IP is used for that subnet. Ensure that Private IP (VPN, AWS Direct Connect, or VPC peering) is used for your chosen subnet if you wish to use this option.

  • You can safely switch between a private connection and a public connection for individual server settings choosing the Use private IP or Do not use private IP option, even after data replication has begun. This switch will only cause a short pause in replication, and will not have any long-term effect on the replication.

  • Choosing the Use private IP option will not create a new private connection.

You should use this option if you want to:

  • Allocate a dedicated bandwidth for replication

  • Use another level of encryption

  • Add another layer of security by transferring the replicated data from one private IP address (source) to another private IP address (on AWS)

The following diagram illustrates the high-level interaction between the different replication system components when using private IP or VPC endpoint.

Application Migration Service network architecture diagram featuring a private link/VPC

Create public IP

When the Use private IP option is chosen, you will have the option to create a public IP. Public IPs are used by default. Choose Create public IP if you want to create a public IP. Choose Do not create a public IP if you do not want to create a public IP.

Dropdown menu showing options to create or not create a public IP address.

Throttle bandwidth

You can control the amount of network bandwidth used for data replication per server. By default, AWS Application Migration Service will use all available network bandwidth utilizing five concurrent connections.

Choose Throttle bandwidth if you want to control the transfer rate of data sent from your source servers to the Replication Servers over TCP Port 1500. Otherwise, choose Do not throttle bandwidth.

Network bandwidth throttling options with radio buttons for throttle and do not throttle.

If you chose to throttle bandwidth, the Throttle network bandwidth (per server, in Mbps) box will appear. Enter your desired bandwidth in Mbps.

Replication resources tags

Add custom Replication resources tags to resources created by AWS Application Migration Service in your AWS account.

These are resources required to facilitate data replication, testing and cutover. Each tag consists of a key and an optional value. You can add a custom tag to all of the AWS resources that are created on your AWS account during the normal operation of AWS Application Migration Service.

To add a new tag, take the following steps:

  1. Click Add new tag.

  2. Enter a Custom tag key and an optional tag value.

Note

Application Migration Service already adds tags to every resource it creates, including service tags and user tags.

These resources include:

  • Amazon EC2 instances

  • Amazon EC2 launch templates

  • Amazon EBS volumes

  • Snapshots

  • Security groups (optional)

Learn more about AWS Tags in this Amazon EC2 article.