Protect and isolate your highly sensitive workloads with a secure enclave
This Guidance shows how you can build a comprehensive cloud architecture for sensitive workloads in national security, defense, and national law enforcement. By using a multi-account architecture on AWS, you can deliver your missions while keeping sensitive data and workloads secure. This Guidance is designed to help you meet strict and unique security and compliance requirements, addressing central identity and access management, governance, data security, comprehensive logging, and network design and segmentation in alignment with various US security frameworks.
Please note: [Disclaimer]
Architecture Diagram
-
Overview
-
Organization Management Account
-
Security Accounts
-
Infrastructure Accounts
-
Application, Community, Team, or Group Accounts (Sensitive)
-
Overview
-
This architecture diagram provides an overview of how to configure comprehensive, multi-account workloads with unique security and compliance requirements. For more details about how to deploy this Guidance, open the other tabs.
Click to enlarge
Step 1
An organization in AWS Organizations with multiple accounts, guided by service control policies (SCPs): The organization groups multiple separate AWS accounts that are controlled by a single customer entity. Separate AWS accounts provide strong control-plane and data-plane isolation between workloads or environments, as if they were owned by different AWS customers.Step 2
The management account is used to create the organization. From the organization’s management account, you can do the following:- Create accounts in the organization and manage policies for all organizational units (OUs).
- Join the following OUs to the organization:
- Security OU
- Infrastructure OU
- Sensitive application OU
Each OU will have one or more member accounts or nested OU, per design.
Step 3
The application OU will have several nested OUs dedicated to application delivery and lifecycle management and will include the following:- Dev OU
- Test OU
- Prod OU
- Shared OU
Additionally, sandbox OUs can also be provisioned as nonsensitive workloads.
Step 1
An organization in AWS Organizations with multiple accounts, guided by service control policies (SCPs): The organization groups multiple separate AWS accounts that are controlled by a single customer entity. Separate AWS accounts provide strong control-plane and data-plane isolation between workloads or environments, as if they were owned by different AWS customers.Step 2
The management account is used to create the organization. From the organization’s management account, you can do the following:- Create accounts in the organization and manage policies for all organizational units (OUs).
- Join the following OUs to the organization:
- Security OU
- Infrastructure OU
- Sensitive application OU
Each OU will have one or more member accounts or nested OU, per design.
Step 3
The application OU will have several nested OUs dedicated to application delivery and lifecycle management and will include the following:- Dev OU
- Test OU
- Prod OU
- Shared OU
Additionally, sandbox OUs can also be provisioned as nonsensitive workloads.
-
Organization Management Account
-
This architecture diagram shows how an organization can group multiple accounts, all controlled by a single customer entity. Follow the steps in this architecture diagram to deploy the Organization Management Account part of this Guidance.
Click to enlarge
Step 1
An organization with multiple accounts: The organization groups multiple separate AWS accounts, which are controlled by a single customer entity. This consolidates billing, groups accounts using OUs, and facilitates the deployment of an organizations preventative controls using SCPs.Step 2
Preventative security controls: These controls, implemented by SCPs, protect the architecture, prevent guardrail disablement, and block undesirable user behavior. SCPs provide a guardrail mechanism principally used to deny specific or entire categories of API operations at an AWS account, OU, or organization level. These can be used to make sure workloads are deployed only in prescribed AWS Regions or deny access to specific AWS services.
Step 3
Automation: Automation makes sure that guardrails are consistently applied when the organization adds new AWS accounts as new teams and workloads are brought onboard. It remediates compliance drift and provides guardrails in the root organization account.
Step 4
Encryption: AWS Key Management Service (AWS KMS) with customer-managed keys encrypts data stored at rest using FIPS 140-2–validated encryption, whether in Amazon Simple Storage Service (Amazon S3) buckets, Amazon Elastic Block Store (Amazon EBS) volumes, Amazon Relational Database Service (Amazon RDS) databases, or other AWS storage services. It protects data in transit using TLS 1.2 or higher.Step 5
Single sign-on: A feature of AWS Identity and Access Management (IAM), IAM Identity Center is used to provide centralized IAM role assumption into AWS accounts across the organization for authorized principals. An organization’s existing identities can be sourced from a customer’s existing Active Directory (AD) identity store or another third-party identity provider (IdP).AWS facilitates multifactor authentication enforcement using authenticator apps, security keys, and built-in authenticators, supporting WebAuthn, FIDO2, and Universal 2nd Factor (U2F) authentication and devices.
Step 1
An organization with multiple accounts: The organization groups multiple separate AWS accounts, which are controlled by a single customer entity. This consolidates billing, groups accounts using OUs, and facilitates the deployment of an organizations preventative controls using SCPs.Step 2
Preventative security controls: These controls, implemented by SCPs, protect the architecture, prevent guardrail disablement, and block undesirable user behavior. SCPs provide a guardrail mechanism principally used to deny specific or entire categories of API operations at an AWS account, OU, or organization level. These can be used to make sure workloads are deployed only in prescribed AWS Regions or deny access to specific AWS services.
Step 3
Automation: Automation makes sure that guardrails are consistently applied when the organization adds new AWS accounts as new teams and workloads are brought onboard. It remediates compliance drift and provides guardrails in the root organization account.
Step 4
Encryption: AWS Key Management Service (AWS KMS) with customer-managed keys encrypts data stored at rest using FIPS 140-2–validated encryption, whether in Amazon Simple Storage Service (Amazon S3) buckets, Amazon Elastic Block Store (Amazon EBS) volumes, Amazon Relational Database Service (Amazon RDS) databases, or other AWS storage services. It protects data in transit using TLS 1.2 or higher.Step 5
Single sign-on: A feature of AWS Identity and Access Management (IAM), IAM Identity Center is used to provide centralized IAM role assumption into AWS accounts across the organization for authorized principals. An organization’s existing identities can be sourced from a customer’s existing Active Directory (AD) identity store or another third-party identity provider (IdP).AWS facilitates multifactor authentication enforcement using authenticator apps, security keys, and built-in authenticators, supporting WebAuthn, FIDO2, and Universal 2nd Factor (U2F) authentication and devices.
-
Security Accounts
-
This architecture diagram shows how to centrally configure a comprehensive log collection across AWS services and accounts. Follow the steps in this architecture diagram to deploy the Security Accounts part of this Guidance.
Click to enlarge
Step 1
Centralized logging: This architecture prescribes comprehensive log collection and centralization across AWS services and accounts. AWS CloudTrail logs work organization-wide to provide full control-plane auditability across the cloud environment.Amazon CloudWatch logs, a cloud-native AWS logging service, is used to capture a wide variety of logs including operating system and application logs, VPC flow logs, and domain name system logs, which are then centralized and only available to defined security personnel.
Step 2
Centralized security monitoring: Compliance drift and security threats are surfaced across the customer’s AWS organization through the automatic deployment of a multitude of different types of detective security controls. This includes activating the multitude of AWS security services in every account in the organization.These security services include Amazon GuardDuty, AWS Security Hub, AWS Config, AWS Firewall Manager, Amazon Macie, IAM Access Analyzer, and CloudWatch alarms. Control and visibility should be delegated across the multi-account environment to a single central security tooling account for easy organization-wide visibility to all security findings and compliance drift.
Step 3
View-only access and searchability: The security account is provided view-only access across the organization (including access to each account’s CloudWatch console) to facilitate investigation during an incident.View-only access is different from read-only access in that it does not provide any access to any data. An optional add-on is available to consume the comprehensive set of centralized logs to make them searchable, providing correlation and basic dashboards.
Step 1
Centralized logging: This architecture prescribes comprehensive log collection and centralization across AWS services and accounts. AWS CloudTrail logs work organization-wide to provide full control-plane auditability across the cloud environment.Amazon CloudWatch logs, a cloud-native AWS logging service, is used to capture a wide variety of logs including operating system and application logs, VPC flow logs, and domain name system logs, which are then centralized and only available to defined security personnel.
Step 2
Centralized security monitoring: Compliance drift and security threats are surfaced across the customer’s AWS organization through the automatic deployment of a multitude of different types of detective security controls. This includes activating the multitude of AWS security services in every account in the organization.These security services include Amazon GuardDuty, AWS Security Hub, AWS Config, AWS Firewall Manager, Amazon Macie, IAM Access Analyzer, and CloudWatch alarms. Control and visibility should be delegated across the multi-account environment to a single central security tooling account for easy organization-wide visibility to all security findings and compliance drift.
Step 3
View-only access and searchability: The security account is provided view-only access across the organization (including access to each account’s CloudWatch console) to facilitate investigation during an incident.
View-only access is different from read-only access in that it does not provide any access to any data. An optional add-on is available to consume the comprehensive set of centralized logs to make them searchable, providing correlation and basic dashboards.
-
Infrastructure Accounts
-
This architecture diagram shows how a centralized, isolated networking environment is built with Virtual Private Clouds (VPCs). Follow the steps in this architecture diagram to deploy the Infrastructure Accounts part of this Guidance.
Click to enlarge
Step 1
Centralized, isolated networking: VPCs built through Amazon Virtual Private Cloud (Amazon VPC) are used to create data-plane isolation between workloads, centralized in a shared-network account. Centralization facilitates strong segregation of duties and cost optimization.Step 2
Mediated connectivity: Connectivity to on-premises environments, internet egress, shared resources, and AWS APIs are mediated at a central point of ingress and egress through the use of AWS Transit Gateway, AWS Site-to-Site VPN, next generation firewalls, and AWS Direct Connect (where applicable).Step 3
Alternative options: The centralized VPC architecture is not for all customers. For customers less concerned with cost optimization, an option exists for local account–based VPCs interconnected through Transit Gateway in the central shared-network account.
Under both options, the architecture prescribes moving AWS public API endpoints into the customer’s private VPC address space, using centralized endpoints for cost efficiency.
Step 4
Centralized ingress and egress infrastructure-as-a-service (IaaS) inspection: It is common to see centralized ingress and egress requirements for IaaS-based workloads. The architecture provides this functionality, so customers can decide if native AWS ingress and egress firewall inspection services—such as AWS Network Firewall, AWS WAF, or Application Load Balancer through Elastic Load Balancing (ELB)—meet their requirements.If not, customers can augment those capabilities with third-party firewall appliances. The architecture supports starting with an AWS firewall and switching to a third-party firewall or using a combination of ingress and egress firewall technologies.
Step 1
Centralized, isolated networking: VPCs built through Amazon Virtual Private Cloud (Amazon VPC) are used to create data-plane isolation between workloads, centralized in a shared-network account. Centralization facilitates strong segregation of duties and cost optimization.Step 2
Mediated connectivity: Connectivity to on-premises environments, internet egress, shared resources, and AWS APIs are mediated at a central point of ingress and egress through the use of AWS Transit Gateway, AWS Site-to-Site VPN, next generation firewalls, and AWS Direct Connect (where applicable).Step 3
Alternative options: The centralized VPC architecture is not for all customers. For customers less concerned with cost optimization, an option exists for local account–based VPCs interconnected through Transit Gateway in the central shared-network account.
Under both options, the architecture prescribes moving AWS public API endpoints into the customer’s private VPC address space, using centralized endpoints for cost efficiency.
Step 4
Centralized ingress and egress infrastructure-as-a-service (IaaS) inspection: It is common to see centralized ingress and egress requirements for IaaS-based workloads. The architecture provides this functionality, so customers can decide if native AWS ingress and egress firewall inspection services—such as AWS Network Firewall, AWS WAF, or Application Load Balancer through Elastic Load Balancing (ELB)—meet their requirements.If not, customers can augment those capabilities with third-party firewall appliances. The architecture supports starting with an AWS firewall and switching to a third-party firewall or using a combination of ingress and egress firewall technologies.
-
Application, Community, Team, or Group Accounts (Sensitive)
-
This architecture diagram shows how to configure segmentation and separation between workloads belonging to different stages of the software development lifecycle, or between different IT administrative roles. Follow the steps in this architecture diagram to deploy the Application, Community, Team, or Group Accounts part of this Guidance.
Click to enlarge
Step 1
Segmentation and separation: The architecture does not merely provide strong segmentation and separation between workloads belonging to different stages of the software development lifecycle or between different IT administrative roles (like between networking, ingress and egress firewalls, and workloads).It also offers a strong network zoning architecture, microsegmenting the environment by wrapping every instance or component in a stateful firewall that is enforced in the hardware of the AWS Nitro System, along with services such as ELB and AWS WAF.
Step 2
All network flows are tightly enforced, with lateral movement prevented between applications, tiers within an application, and nodes in a tier of an application unless explicitly allowed. Further, routing is prevented between Dev, Test, and Prod with recommendations on a CI/CD architecture to enable developer agility and ease code promotion between environments with appropriate approvals.
Step 1
Segmentation and separation: The architecture does not merely provide strong segmentation and separation between workloads belonging to different stages of the software development lifecycle or between different IT administrative roles (like between networking, ingress and egress firewalls, and workloads).It also offers a strong network zoning architecture, microsegmenting the environment by wrapping every instance or component in a stateful firewall that is enforced in the hardware of the AWS Nitro System, along with services such as ELB and AWS WAF.
Step 2
All network flows are tightly enforced, with lateral movement prevented between applications, tiers within an application, and nodes in a tier of an application unless explicitly allowed. Further, routing is prevented between Dev, Test, and Prod with recommendations on a CI/CD architecture to enable developer agility and ease code promotion between environments with appropriate approvals.
Well-Architected Pillars
The AWS Well-Architected Framework helps you understand the pros and cons of the decisions you make when building systems in the cloud. The six pillars of the Framework allow you to learn architectural best practices for designing and operating reliable, secure, efficient, cost-effective, and sustainable systems. Using the AWS Well-Architected Tool, available at no charge in the AWS Management Console, you can review your workloads against these best practices by answering a set of questions for each pillar.
The architecture diagram above is an example of a Solution created with Well-Architected best practices in mind. To be fully Well-Architected, you should follow as many Well-Architected best practices as possible.
-
Operational ExcellenceRead the Operational Excellence whitepaper
This Guidance uses Organizations with AWS CloudFormation stacks and configurations to create a secure foundation for your AWS environment. This provides an infrastructure-as-code (IaC) solution that accelerates your implementation of technical security controls. Config rules remediate any configuration deltas that have been determined to negatively impact the prescribed architecture. You can use the AWS global commercial infrastructure for sensitive classified workloads and automate secure systems to deliver missions faster while continually improving your processes and procedures.
-
SecurityRead the Security whitepaper
This Guidance uses Organizations to facilitate the deployment of organizational guardrails, such as API logging with CloudTrail. This Guidance also provides preventative controls using prescriptive AWS SCPs as a guardrail mechanism, principally used to deny specific or entire categories of APIs within your environment (to make sure that workloads are deployed only in prescribed Regions) or deny access to specific AWS services. CloudTrail and CloudWatch logs support a prescribed comprehensive log collection and centralization across AWS services and accounts. AWS security capabilities and the multitude of security-relevant services are configured in a defined pattern that helps you meet some of the strictest security requirements in the world.
-
ReliabilityRead the Reliability whitepaper
This Guidance uses multiple Availability Zones (AZs), so the loss of one AZ does not impact application availability. You can use CloudFormation to automate the provisioning and updating of your infrastructure in a safe and controlled manner. This Guidance also provides prebuilt rules for evaluating AWS resource configurations and configuration changes within your environment, or you can create custom rules in AWS Lambda to define best practices and guidelines. You can automate the ability to scale your environment to meet demand and mitigate disruptions such as misconfigurations or transient network issues.
-
Performance EfficiencyRead the Performance Efficiency whitepaper
This Guidance simplifies cloud infrastructure management by using Transit Gateway, which serves as a central hub that connects multiple VPCs through a single gateway, making it easier to scale and maintain the network architecture. This simplifies your network architecture and facilitates efficient traffic routing between different AWS accounts within your organization.
-
Cost OptimizationRead the Cost Optimization whitepaper
This Guidance provides the ability to avoid or remove unneeded costs or the use of suboptimal resources. Organizations provides centralization and consolidated billing, facilitating the strong separation of resource use and cost optimization. This Guidance prescribes moving AWS public API endpoints into your private VPC address space, using centralized endpoints for cost efficiency. Additionally, you can use AWS Cost and Usage Reports (AWS CUR) to track your AWS usage and estimate charges.
-
SustainabilityRead the Sustainability whitepaper
This Guidance helps you reduce the carbon footprint associated with managing workloads within your own datacenters. The AWS global infrastructure offers supporting infrastructure (such as power, cooling, and networking), a higher utilization rate, and faster technology refreshes than traditional data centers. Additionally, the segmentation and separation of workloads helps you reduce unnecessary data movement, and Amazon S3 offers storage tiers and the ability to automatically move data to efficient storage tiers.
Implementation Resources
The sample code is a starting point. It is industry validated, prescriptive but not definitive, and a peek under the hood to help you begin.
Related Content
TSE-SE Sample Configuration (with LZA automation engine)
Trusted Secure Enclaves - Sensitive Edition
Disclaimer
The sample code; software libraries; command line tools; proofs of concept; templates; or other related technology (including any of the foregoing that are provided by our personnel) is provided to you as AWS Content under the AWS Customer Agreement, or the relevant written agreement between you and AWS (whichever applies). You should not use this AWS Content in your production accounts, or on production or other critical data. You are responsible for testing, securing, and optimizing the AWS Content, such as sample code, as appropriate for production grade use based on your specific quality control practices and standards. Deploying AWS Content may incur AWS charges for creating or using AWS chargeable resources, such as running Amazon EC2 instances or using Amazon S3 storage.
References to third-party services or organizations in this Guidance do not imply an endorsement, sponsorship, or affiliation between Amazon or AWS and the third party. Guidance from AWS is a technical starting point, and you can customize your integration with third-party services when you deploy the architecture.