There are two types of AWS accounts that you can offboard from AMS Advanced multi-account landing zone:
-
Application accounts
-
Core accounts
To offboard all accounts from your AMS Multi-account landing zone, you must offboard all Application accounts before you offboard Core accounts.
To take over and continue operating workloads in offboarded Application or Core accounts, make sure that you review this documentation with your AMS account team. This documentation outlines the changes that AMS performs during the offboard process.
Tasks to complete for continued operation of offboarded accounts
The following tasks are required for continued operation of accounts that you offboarded from AMS multi-account landing zone:
Turn on Developer mode: To gain more permission to your accounts, turn on Developer mode before you offboard Application accounts from AMS. When you turn on Developer mode, you can more easily make the necessary changes to prepare for offboarding. Don’t try to remove or modify AMS infrastructure resources. If you delete AMS infrastructure resources, then AMS might not be able to successfully offboard your account. For information about how to enable Developer mode, see Getting started with AMS Advanced Developer mode.
If you can’t complete the necessary changes to prepare for offboarding after you turn on Developer mode, then contact your AMS account team to discuss your requirements.
Choose an alternate method for EC2 stack access: After you offboard Application accounts from AMS, you can’t use RFCs to access your stack resources. Review Offboarding changes, and then choose an alternate access method so that you retain access to your stacks. For more information, see Access Alternatives.
Offboard AMS Application accounts
To offboard Application accounts from your multi-account landing zone environment, complete the following steps for each account:
Verify that there are no open RFCs in the account. For more information, see Create, clone, update, find, and cancel RFCs.
Verify that you can access the primary or root user email address for the account.
From the Application account, submit an RFC with the Application Account | Confirm offboarding (ct-2wlfo2jxj2rkj) change type. In the RFC, specify the Application account to offboard.
From the Management account, submit an RFC with the Management account | Offboard Application Account (ct-0vdiy51oyrhhm) change type. In the RFC, specify the Application account to offboard. Also, indicate if you want to delete or retain the transit gateway attachment to the landing zone.
To make sure that AMS billing is stopped, notify your CSDM that you offboarded the account.
The following occurs after the Application account is offboarded:
All components are disassociated from AMS services, but your created resources remain in the account. You can choose to keep or close the AMS offboarded account.
Core accounts and other remaining Application accounts function normally after an Application account is offboarded.
AMS billing is stopped, but AWS billing isn’t stopped until you close the account. For more information, see What you need to know before closing your account.
If an account is closed, then the account is visible in your organization in the
suspended
state for 90 days. After 90 days, the closed account is permanently removed and no longer visible in your organization.After the account is closed, you can still sign in and file a support case or contact Support for 90 days.
After 90 days, any content that remains in the account is permanently deleted and the remaining AWS services are terminated.
- Q: Can I use my federated IAM roles to continue to access an Application account that I offboarded from my AMS multi-account landing zone?
Yes. AMS created default AWS Identity and Access Management (IAM) roles remain available in the account after AMS offboarding. However, these roles and policies are designed for use with AMS access management. To provide the necessary access for your users, you might need to deploy your own IAM resources.
- Q: How do I gain full access to an Application account that I offboarded from my AMS multi-account landing zone?
Offboarded Application accounts are moved to the Deprecated Organizational Unit (OU) in the AWS Organizations account structure. This move lifts SCP access restrictions that previously blocked root user access. For information about how to reset root user credentials, see Resetting a lost or forgotten root user password.
- Q: What changes are made during Application account offboarding?
For information about actions that AMS takes when the service offboards accounts, see Offboarding changes.
- Q: Can I offboard an Application account without detaching it from the transit gateway?
Yes. Use the Management account | Offboard Application Account (ct-0vdiy51oyrhhm) change type to submit the RFC, and specify the
DeleteTransitGatewayAttachment
parameter asFalse
.- Q: How long does it take to offboard an Application account?
When you use the Management account | Offboard Application Account (ct-0vdiy51oyrhhm) change type, RFCs complete within 1 hour.
- Q: Is it mandatory that I close the offboarded account?
No. Account closure after AMS offboarding isn’t mandatory. During the offboarding process, AMS removes its access and management of your AWS account, but your account and your resources within the account remain. It's important to note that after AMS offboarding, you’re solely responsible for managing and maintaining your AWS account and resources. AMS isn’t responsible for any issues, incidents, or service disruptions that might occur in your account after the offboarding process is complete. For more information, see How do I close my AWS account?
. - Q: If I submit an account closure request, are all existing resources deleted immediately?
No. Account closure doesn’t terminate your resources. Resources in the account automatically terminate 90 days after the closure request. AMS billing stops, but AWS resource billing doesn’t stop until you close the account. For more information, see What you need to know before closing your account.
- Q: Can I schedule the offboarding of an Application account?
Yes. You can schedule the RFCs to run at a specific time. However, the Application Account | Confirm offboarding RFC must complete before you can schedule the Management account | Offboard Application Account RFC. For more information, see RFC scheduling.
-
R: Responsible party. The party responsible for completing the listed task.
-
A: Accountable party. The party that approves the completed task.
-
C: Consulted party. A party whose opinions are sought, typically as subject matter experts, and with whom there is bilateral communication.
-
I: Informed party. A party that’s informed on progress, often only on completion of the task or deliverable.
Activity | Customer | AWS Managed Services (AMS) |
---|---|---|
Prerequisites | ||
Verify access to the root email address for each AWS account ID that will be offboarded | R | C |
Review AMS documentation on recommended customer actions and prepare accounts for AMS offboarding | R | C |
If needed, submit an RFC to enable Developer mode to prepare accounts for AMS offboarding | R | I |
If needed, choose an alternate method for EC2 stack access. | R | I |
Offboarding | ||
Submit RFCs to confirm and request offboarding of Application accounts | R | I |
Offboard AMS components from Application accounts | I | R |
Notify AMS CSDM of offboarded accounts to stop AMS billing | R | I |
Post-offboarding | ||
Reset root user account password and verify root access in offboarded accounts | R | C |
Close the account or follow AMS guidance on recommended customer actions in the AMS offboarding documentation to continue operating the accounts | R | C |
Offboard Core accounts
To offboard multi-account landing zone Core accounts, complete the following steps:
Verify that all Application accounts in the landing zone were offboarded from AMS.
Verify that you have no open RFCs in the accounts. For more information, see Create, clone, update, find, and cancel RFCs.
Verify that you can access the primary or root user email address for all Core accounts. For more information, see Multi-Account Landing Zone accounts.
Verify that you can access the primary or root user phone number for the management account. Use the
AWSManagedServicesBillingRole
IAM role to update the phone number. For more information, see How do I update my telephone number that's associated with my AWS account?. Log in to your AMS landing zone management account and submit an AMS service request. In the service request, specify to offboard your entire landing zone.
The following occurs after the Core accounts are offboarded:
-
All components are disassociated from AMS services, but some AWS resources remain in the account. You can choose to keep or close the AMS offboarded Core accounts.
-
AMS billing is stopped, but AWS billing isn’t stopped until you close the account. For more information, see What you need to know before closing your account.
-
If an account is closed, then the account is visible in your organization in the
suspended
state for 90 days. After 90 days, the closed member account is permanently removed and is no longer visible in your organization. -
After the account is closed, you can still sign in and file a support case or contact Support for 90 days.
-
After the account is closed for 90 days, any content that remains in the account is permanently deleted, and the remaining AWS services are terminated.
- Q: Can I use my federated IAM roles to continue to access the offboarded Core accounts?
Yes. AMS created default AWS Identity and Access Management (IAM) roles remain available in the offboarded account. However, these roles and policies are designed for use with AMS access management. To provide the necessary access for your users, you might need to deploy your own IAM resources.
- Q: How do I gain full access to the Management, Shared Services, Networking, or other non-Application MALZ account after offboarding from AMS multi-account landing zone?
After offboarding, follow the instructions in Resetting a lost or forgotten root user password to use primary (root) user credentials to access Core accounts other than the management account. Unlike other account types, the management account retains an inaccessible multi-factor authentication (MFA) device that’s associated to the root user to prevent usage. To regain root access you must follow the lost MFA device recovery process
. - Q: What changes are made during Core account offboarding?
For information about actions that AMS takes when the service offboards accounts, see Offboarding changes.
- Q: How long does Core account offboarding take to complete?
The Core account offboarding process typically takes up to 30 days to complete. However, to make sure that all required steps are correctly completed, you must initiate the offboarding request at least 7 days before offboarding starts. To facilitate an easy transition, plan ahead and submit your offboarding request in advance.
- Q: How do I manage shared components after AMS offboarding?
AMS Managed Active Directory and other shared services infrastructure components are designed for AMS operator access. You might need to update Amazon Elastic Compute Cloud (Amazon EC2) security groups, AWS Organizations service control policy (SCP), or make other changes to retain full access to these components.
- Q: Can I close offboarded Core accounts?
By default, Application accounts have multiple dependencies on MALZ Core accounts, such as AWS Organizations membership, transit gateway network connectivity, and DNS resolution through AMS Managed Active Directory. After you resolve these dependences, you can decommission and close the offboarded Core account. For more information, see Multi-Account Landing Zone accounts.
-
R: Responsible party. The party responsible for completing the listed task.
-
A: Accountable party. The party that approves the completed task.
-
C: Consulted party. A party whose opinions are sought, typically as subject matter experts, and with whom there is bilateral communication.
-
I: Informed party. A party that’s informed on progress, often only on completion of the task or deliverable.
Activity | Customer | AWS Managed Services (AMS) |
---|---|---|
Prerequisites | ||
Verify access to the root email address for each AWS account ID that will be offboarded | R | C |
Verify access to and update the root user phone number for the management account | R | C |
Review AMS documentation on recommended customer actions and prepare accounts for AMS offboarding | R | C |
Offboarding | ||
Submit service request to request offboarding of the landing zone | R | I |
Offboard AMS components from Core accounts | I | R |
Post-offboarding | ||
Reset root user account password and verify root access in offboarded accounts | R | C |
Close the accounts or follow AMS guidance on recommended customer actions in AMS offboarding documentation to continue to operate the accounts | R | C |
Offboarding changes
The following table describes the actions that AMS takes for multi-account landing zone offboarding, potential impacts, and recommended actions.
Component | Account type | Actions taken to offboard | Potential impacts | Recommended customer action |
---|---|---|---|---|
Access management | Application accounts |
After offboarding, stack access RFCs for just-in-time, time-bound access can no longer be submitted to access EC2 stacks through AMS bastion hosts AMS no longer manages access-related components on any existing EC2 resource stacks (PBIS Open agent, domain join scripts) |
Can’t use AMS bastions through RFS to access EC2 instances EC2 instances launched from non AMS provided AMIs aren’t joined to Managed Active Directory domain If not removed, AMS launch scripts in existing resource stacks might produce errors because of missing AMS dependencies, and prevent rejoining to a different domain |
Use alternate methods to access EC2 instance (see Access Alternatives) Remove AMS launch scripts from existing EC2 resource stacks (see Disable AMS EC2 launch scripts) |
Access management (cont’d) | Core accounts | If you migrated from PBIS Open to PBIS Enterprise (AD Bridge), then AMS no longer renews licensing after core account offboarding | If PBIS Enterprise license is allowed to expire, Active Directory credentials aren’t valid for existing Linux-based EC2 instance stacks | If you migrated to PBIS Enterprise (AD Bridge), then decide whether to maintain licensing or decommission (see PBIS Open/Enterprise (AD Bridge)) |
Logging, Monitoring, Incident/Event Management | Application and Core accounts |
AMS components to deploy Alerts from baseline monitoring in AMS are removed Existing deployed Amazon CloudWatch alarms remain but no longer create AMS incidents AWS Config aggregation authorizations from AMS and the MALZ Core security account are removed AWS Config rules remain deployed and Amazon GuardDuty remains enabled, but no longer creates AMS incidents |
Newly created resources don’t have AMS baseline monitoring and alarms applied Infrastructure metric alarms and security events no longer generate AMS incidents AWS Config is no longer aggregated in a central account |
Define, capture, and analyze operations metrics to view workload events and take appropriate action. Implement any required alert workflow to continue to apply required operational monitoring and alarms on new resources and to for receive security alerts from AWS Config and Amazon GuardDuty. |
Continuity management (Backup and Restore) | Application accounts |
AMS no longer monitors backup jobs or performs backup and restore requests AMS default backup vaults, backup encryption key, and backup role remain |
Backup operation failures might not be noticed | Monitor and review backup plan configurations |
Patch management | Application and Core accounts |
AMS no longer monitors patching operations for successful execution or performs manual patching AMS no longer updates AMS infrastructure components AMS provided patch baselines are retained AMS provided AWS Systems Manager patch automation runbooks are unshared and no longer available for use |
Patching operation failures might not be noticed Existing patch configurations that depend on AMS provided Systems Manager Automation runbooks must be reconfigured to continue uninterrupted |
Review and reconfigure patching configurations as needed |
Network management | Application accounts | If specified, the transit gateway attachment in the offboarded Application account is removed | Offboarded Application account can no longer use a transit gateway to access shared services, such as Managed Active Directory or other Application accounts |
Specify DeleteTransitGatewayAttachment as False to retain
the transit gateway connectivity
|
Security management | Application accounts |
Account is detached from central Trend Micro DSM console. Also, endpoint agents no longer forward alerts through the AMS incident process Trend Micro agents remain installed but are no longer managed or updated by AMS AMS provided AMI customizations that deploy Trend Micro agent are no longer maintained or updated by AMS |
EC2 instance endpoint malware detections might not be noticed The Trend micro agent isn’t deployed on EC2 instances that are launched from non AMS provided AMIs |
Consider options for continuing or discontinuing Trend Micro (see Trend Micro Deep Security) |
Security management (cont’d) | Core accounts |
Trend Micro DSM infrastructure is left in place within Shared Services accounts but no longer maintained or updated by AMS Trend Micro DSM no longer forwards alerts through the AMS incident process |
EC2 instance endpoint malware detections might go unnoticed EC2 instance endpoint protection might be impacted if infrastructure isn’t maintained (definition updates, licensing, and so on) |
Decide whether to continue or discontinue Trend Micro (see (see Trend Micro Deep Security) |
Change management | Application and Core accounts |
AMS RFC console and API is removed AMS custom service control policies (SCPs) that contain account-level access restrictions are detached during Application account offboarding, and deleted during Core account offboarding |
You must use a native AWS API to create new resources, change existing resources, or update existing AWS CloudFormation stacks Access restrictions are no longer imposed at the account-level through AMS provided SCPs |
Make sure that user roles provide sufficient access to use AWS services Create SCPs to provide account-level permission restrictions |
AMS OS images and automations for service management | Application and Core accounts |
AMS no longer provides support on customizations and launch scripts included in AMS provided EC2 AMIs AMS provided EC2 AMIs remain available in your offboarded accounts AMS provided Systems Manager Automation runbooks are unshared and no longer available for use |
After offboarding, AMS provided AMIs launched with AWS CloudFormation send cfn-signal Operational processes that depend on AMS provided Systems Manager Automation runbooks might fail |
Review and update any build or operational processes that are dependent on AMS provided AMIs or Systems Manager Automation runbooks |
Shared services infrastructure | Core accounts | AMS access is removed and AMS no longer manages shared components, including AMS Managed Active Directory, AWS Transit Gateway, and AWS Organizations | Loss of management over shared infrastructure | Reset admin access to AMS Managed Active Directory and assume management of shared services components |
Reporting | Application and Core accounts | AMS no longer collects account or resource-level details for aggregate reporting | Loss of insight into operational and business metrics (backup and patching coverage, change management, and incident activity) | Replace any needed aggregate data reporting across accounts with their own solution |
AMS account team and service desk | Application and Core accounts | AMS account team (CSDM, CA) and AMS operations service desk no longer support the offboarded accounts | Loss of operational support with expertise in the AMS designed multi-account landing zone architecture and related components | Make sure that there’s sufficient personnel and familiarity with account structure and resources to support operations in the environment |
Access Alternatives
The following are alternatives methods for retaining access to your EC2 stack after you offboard AMS accounts:
Use Session Manager to access EC2 instances with elevated permissions without requiring bastions or inbound network access. For more information, see AWS Systems Manager Session Manager.
Rejoin EC2 instances to a different Active Directory domain with new domain credentials. If you use AWS Directory Service, then see Join an EC2 instance to your AWS Managed Microsoft AD directory.
Use local user accounts that you created through one of the other access methods or through AWS Systems Manager Run Command.
Disable AMS EC2 launch scripts
Linux operating systems
Use your distribution’s package manager to uninstall the
ams-modules
package. For example, for Amazon Linux 2 use yum remove ams-modules
.
Windows operating systems
To disable EC2 launch scripts in Windows, complete the following steps:
Windows Server 2008/2012/2012r2/2016/2019:
Disable or remove the Managed Services Startup scheduled task from Task Scheduler. To list scheduled tasks, run the
Get-ScheduledTask -TaskName '*Ec2*'
command.Windows Server 2022:
Remove the EC2Launch v2 task. This task runs
Initialize-AMSBoot
inpostReady
stage in C:\ProgramData\Amazon\EC2Launch\config\agent-config.yml on the instance. The following is a snippet from an exampleagent-config.yml
:{ "task": "executeScript", "inputs": [ { "frequency": "always", "type": "powershell", "runAs": "localSystem" } ] }
(Optional) Remove the following file contents:
C:\Program Files\WindowsPowerShell\Modules\AWSManagedServices.* C:\Windows\System32\WindowsPowerShell\v1.0\Modules\AWSManagedServices.Build.Utilities\*
PBIS Open/Enterprise (AD Bridge)
To determine if you use PBIS Open or PBIS Enterprise (AD Bridge) edition, run the following command in a Linux EC2 managed instance:
yum info | grep pbis
The following is example output that shows PBIS Enterprise (AD Bridge):
Name : pbis-enterprise
From repo : pbise
Name : pbis-enterprise-devel
Repo : pbise
Description : The pbis-enterprise-devel package includes the development
PBIS Open
PBIS Open is a deprecated product that BeyondTrust no longer supports. For more information, see the BeyondTrust pbis-open documentation
AD Bridge (PBIS Enterprise)
You can do one of the following:
Renew licensing and continue operating AD Bridge. Contact BeyondTrust to discuss licensing and support.
Discontinue use of AD Bridge. Run the following Shell command to remove PBIS-Enterprise package from Linux managed instances. For more information, see the BeyondTrust documentation Leave a Domain and Uninstall the AD Bridge Agent
. $ sudo /opt/pbis/bin/uninstall.sh purge
Leave the AMS managed Active Directory domain without removing PBIS agent
You have the option to leave the AMS managed Active Directory without removing the PBIS agent. Use one of the following solutions, depending on your operating system:
Use PBIS from the AMS managed AD to run the following shell command to unjoin a Linux EC2 instance. For more information, see the BeyondTrust pbis-open
$ sudo /opt/pbis/bin/domainjoin-cli leave
You might see an error message similar to the following:
Error: LW_ERROR_KRB5_REALM_CANT_RESOLVE [code 0x0000a3e1]
Cannot resolve network address for KDC in requested realm
If this error occurs, then run the following commands to delete AD provider registry and restart lwsm
services:
$ /opt/pbis/bin/regshell dir '[HKEY_THIS_MACHINE\Services\lsass\Parameters\Providers\ActiveDirectory\DomainJoin]'
Use the directory ID output that you received from the previous command (for example, A123EXAMPLE.AMAZONAWS.COM
) to run the following commands:
$ /opt/pbis/bin/regshell delete_tree \ '[HKEY_THIS_MACHINE\Services\lsass\Parameters\Providers\ActiveDirectory\DomainJoin\
DIRECTORYID
]' $ /etc/pbis/redhat/lwsmd restart $ /opt/pbis/bin/lwsm restart lwreg
Trend Micro Deep Security
Use one of the following options to continue or discontinue the use of Trend Micro Deep Security:
Continue usage
(If offboarding the entire MALZ) After Core account offboarding, reconnect offboarded Application accounts to the existing Trend Micro Deep Security Manager (DSM) and maintain licensing in shared Services account. For more information, see Add AWS cloud accounts
and Check your license information . Log in to the shared services account and navigate to the Secrets Manager console.
Retrieve DSM console admin credentials that are stored in the
/ams/eps/
path.Log in to the DSM console at https://dsm.sentinel.int
. Choose Use Cross Account Role, and then enter
arn:aws:iam::ACCOUNTID:role/mc_eps_cross_account_role
. Replace ACCOUNTID with the offboarded Application account ID.Choose Next.
Wait several minutes for DSM to process the account discovery and show that sync was successful.
Reconnect offboarded Application accounts to a new Trend Micro DSM installation. For more information, see Activate and protect agents
and Activate the agent . Reconnect offboarded Application accounts to Trend Micro Cloud One. For more information, see Migrate from Deep Security to Workload Security
and Migrate from an on-premises DSM .
Discontinue usage
Uninstall Trend Micro Deep Security Agents from offboarded Application accounts. For more information, see Uninstall Deep Security
.