Skip to content

Landing Zone

Last reviewed: August 2026

A landing zone is the foundational setup for operating a multi-account cloud environment securely and consistently. It generally includes the following elements:

  • Network: standardized VPC/VNet structure, hub-and-spoke topology, connectivity policies
  • Security: security boundaries between accounts, encryption policies, threat detection
  • Logging: centralized log collection, audit trails, compliance evidence
  • Guardrails: preventive/detective policies applying consistent governance across the organization

A landing zone lets you provision new workload accounts quickly and safely, while automatically applying your organization’s security and compliance requirements.

graph TB
    subgraph "Landing Zone Components"
        A[Organization Root] --> B[Security OU]
        A --> C[Shared Services OU]
        A --> D[Workload OU]
        B --> B1[Log Archive Account]
        B --> B2[Security Audit Account]
        C --> C1[Network Hub Account]
        C --> C2[Shared Services Account]
        D --> D1[Production Account]
        D --> D2[Development Account]
        D --> D3[Test Account]
    end
Item AWS Azure Google Cloud OCI
Service name AWS Control Tower Azure Landing Zone Google Cloud Foundation Toolkit OCI Landing Zone
Account structure AWS Organizations + OU Management Group + Subscription Organization + Folder + Project Tenancy + Compartment
Guardrails Controls (preventive/detective/proactive) Azure Policy + Deployment Stacks Organization Policy CIS Benchmark-based policies
Default network structure VPC + Transit Gateway Hub-Spoke VNet + Azure Firewall Shared VPC + Cloud Interconnect Hub-Spoke VCN + DRG
IaC support AWS CloudFormation (built-in) Bicep / Terraform modules Terraform modules Terraform modules
Logging AWS CloudTrail + Config Activity Log + Defender for Cloud Cloud Audit Logs + Security Command Center Audit + Cloud Guard

When designing a landing zone, first work out the following items.

  • Organization structure — decide the criteria for dividing accounts, subscriptions, projects, and compartments.
  • Environment separation — separate production, development, test, security, and shared-services accounts.
  • Network boundaries — decide the hub-and-spoke design, internet entry points, and on-premises connectivity method.
  • Identity integration — standardize your in-house IdP, SSO, MFA, and how admin privileges are granted.
  • Logging and audit — collect audit logs from all accounts into a central repository.
  • Guardrails — automatically apply policies such as restricted regions, blocking public storage, and enforced encryption.
Stage Description
1 First build a minimal account structure and a central log repository.
2 Template network, security, and IAM standards as IaC.
3 Automate the procedure for creating new workload accounts.
4 Operate policy violation detection and an exception approval process.

VPC design is a core element of a landing zone. Separate VPCs per workload to create security boundaries, and place shared services in a central VPC.

Separation Criteria Example Reason
By environment dev / staging / prod VPC Production isolation, mistake prevention
By team/service Team A VPC / Team B VPC Security boundaries, independent operation
By regulation PCI VPC / general VPC Minimizing compliance scope

A pattern where logging, security tools, DNS, proxies, and similar services are placed in a central VPC and accessed from other VPCs.

Vendor Account/project separation VPC sharing Hub connectivity
AWS Organizations + OU Subnet sharing via RAM Transit Gateway
Azure Separation by subscription Hub-Spoke VNet Virtual WAN
Google Cloud Shared VPC (host + service projects) Subnet sharing from the host project VPC Peering / NCC
OCI Compartment separation DRG hub

Allocate CIDR blocks in advance to prevent IP range conflicts between accounts/projects.

  • Reserve a /8 or /10 range for the entire organization, and distribute /16/20 units per account/environment
  • If CIDR ranges overlap during VPC peering/Transit Gateway connections, routing becomes impossible
  • Allocate generously to account for future growth (adding subnets, creating new accounts)

For detailed network design, see VPC and Subnets.

  • Have you completed the organization structure design (OU/folder/compartment)?
  • Have you decided on an environment separation strategy (production/staging/development/security/shared services)?
  • Have you decided on a network topology (hub-and-spoke, centralized egress)?
  • Have you defined guardrails (preventive/detective policies)?
  • Have you configured a central logging account (CloudTrail/Activity Log aggregation)?
  • Have you separated a security audit account?
  • Have you integrated an identity provider (IdP) (SSO/SAML)?
  • Have you defined a cost allocation tagging policy?
  • Have you set up an automated pipeline for provisioning new accounts (IaC)?
  • Have you documented a break-glass (emergency access) procedure?
  • Creating VPCs without a CIDR plan — later, VPC peering/Transit Gateway connections become impossible to route due to IP range conflicts
  • Creating workload accounts before central logging is in place — audit logs end up scattered across accounts, making integrated investigation impossible during a security incident
  • Deploying accounts without guardrails — developers accidentally create public S3 buckets or provision resources in restricted regions

AWS Control Tower Landing Zone 4.0 (November 2025) moved away from a “single enforced blueprint” approach toward a modular model where you apply only the controls and choose the rest.

Before After LZ 4.0
Setting up Control Tower created a fixed OU/account structure Controls can be applied while keeping your existing organizational structure
All service integrations were provided as a package Service integrations (Config, CloudTrail, etc.) can be enabled selectively
Difficult to customize in large organizations Can run alongside existing IaC pipelines

Azure’s Cloud Adoption Framework (CAF) also continues to expand its modular landing zone architecture, and Google Cloud Foundation Toolkit supports similar selective application based on Terraform modules.

As data sovereignty requirements intensify, landing zones have emerged that keep not just data storage but also processing within the jurisdiction.

Vendor Solution Key Features Timing
Microsoft Cloud for Sovereignty — SLZ + Sovereign Public Cloud Data residency guardrails, confidential computing, EU Data Boundary, Data Guardian, IaC policies 2025-2026
Google Cloud Sovereign Cloud Controls In-jurisdiction processing, key management, access transparency 2025-2026
AWS Sovereign Controls (Control Tower + Nitro) Region-restriction guardrails, Nitro confidential computing, data residency policies Existing (ongoing enhancement)
OCI EU Sovereign Cloud Physically isolated EU-only infrastructure Existing
Regulation Impact on Landing Zone
DORA (effective 2025.01.17) Requires financial institutions to manage ICT third-party risk → cloud vendors must be managed as critical ICT providers, exit strategy required
NIS2 Strengthens security obligations for critical infrastructure operators → requires landing-zone-level governance evidence
EU AI Act (GPAI obligations 2025.08; sanction powers and Article 50 transparency 2026.08; high-risk AI deferred by Digital Omnibus — standalone 2027.12, product-embedded 2028.08) Data governance requirements for high-risk AI systems → landing zone must include data classification/access control per AI workload