7.4 Technology Architecture Domain: Hardware, Software, Cloud & Network Platforms
Key Takeaways
- Technology Architecture (Phase D) defines the physical and virtual compute, storage, networking, and platform software infrastructure required to host Application and Data components.
- Platform Services provide runtime execution environments, container orchestrators, database engines, and middleware, abstracting physical hardware from application software.
- Cloud Service Models (IaaS, PaaS, SaaS) reallocate operational responsibilities between enterprise architecture teams and cloud service providers across hybrid and multi-cloud environments.
- The Technology Standards Catalog enforces standardized hardware and software stacks aligned with the Technical Reference Model (TRM) to manage technical debt and prevent vendor lock-in.
- Technology Lifecycle Management utilizes N-1 support strategies and proactive obsolescence tracking to retire legacy infrastructure before security and operational risks materialize.
7.4 Technology Architecture Domain: Hardware, Software, Cloud & Network Platforms
Technology Architecture represents Phase D of the TOGAF Architecture Development Method (ADM). It defines the software, platform, network, and hardware infrastructure required to support the deployment and execution of application services, software components, and data repositories defined in Phase C.
Technology Architecture provides the physical and virtual foundation of the enterprise architecture. Without a resilient, scalable Technology Architecture, higher-level application components and business processes will experience performance bottlenecks, security vulnerabilities, and operational downtime.
Platform Services vs. Infrastructure Components
TOGAF Technology Architecture categorizes infrastructure elements into two distinct categories: Platform Services and Infrastructure Components.
+---------------------------------------------------------------------------------+
| APPLICATION & DATA ARCHITECTURE (Phase C Deliverables) |
+---------------------------------------------------------------------------------+
|
v Hosted On
+---------------------------------------------------------------------------------+
| PLATFORM SERVICES (Technical Runtime Middleware) |
| OS (Linux/Windows) | Runtime (.NET/JVM) | Container Orchestration (Kubernetes) |
| DBMS Engine (PostgreSQL) | Message Broker (Kafka) | Identity Provider (Okta) |
+---------------------------------------------------------------------------------+
|
v Executed By
+---------------------------------------------------------------------------------+
| INFRASTRUCTURE COMPONENTS (Physical & Virtual Hardware Assets) |
| Compute (x86/ARM Servers, VMs) | Storage (SAN, NAS, NVMe) |
| Network (Switches, Routers, Firewalls) | Cloud Data Center Facilities (AWS, Azure) |
+---------------------------------------------------------------------------------+
1. Platform Services (Technical Middleware & Runtimes)
Platform Services are technical software building blocks that provide execution runtimes, system utilities, and middleware infrastructure to support software applications. They abstract underlying physical hardware capabilities for application developers.
- Operating Systems: Linux (RHEL, Ubuntu), Windows Server.
- Container Orchestrators: Kubernetes, Docker Swarm, OpenShift.
- Application Runtimes: Java Virtual Machine (JVM), Node.js, .NET Core runtime.
- Database Management Engines: PostgreSQL, Oracle DB, Cassandra, Redis.
- Integration & Messaging Middleware: Apache Kafka, RabbitMQ, IBM MQ.
2. Infrastructure Components (Hardware & Network Assets)
Infrastructure Components represent the physical hardware, virtualized hardware resources, network devices, and data center facilities that host platform services.
- Compute Resources: Bare-metal x86/ARM servers, hypervisors (VMware ESXi, KVM), virtual machines.
- Storage Infrastructure: Storage Area Networks (SAN), Network Attached Storage (NAS), NVMe arrays, cloud S3 storage buckets.
- Network & Security Topology: Physical switches, routers, software-defined networking (SDN), firewalls, load balancers (F5, AWS ALB), VPN gateways, and CDN edge nodes.
Cloud Computing Service Models & Deployment Architectures
Modern Technology Architecture incorporates Cloud Computing Service Models that redefine boundary responsibilities between internal enterprise IT and external cloud service providers.
| Cloud Model | Customer Responsibility Scope | Provider Responsibility Scope | Architectural Use Case |
|---|---|---|---|
| IaaS (Infrastructure as a Service) | OS, Runtimes, Middleware, Data, Applications | Physical Hardware, Hypervisors, Storage, Data Center Facilities | Lift-and-shift legacy migrations; custom network control. |
| PaaS (Platform as a Service) | Data, Application Code, Configurations | OS, Runtimes, Middleware, Database Engines, Hardware | Cloud-native application development; rapid DevOps deployment. |
| SaaS (Software as a Service) | User Access Management, Data Entry | Entire Application, Runtimes, Infrastructure, Facilities | Standard business support tools (Salesforce, Office 365, Workday). |
Hybrid Cloud and Multi-Cloud Topologies
Enterprises increasingly deploy Hybrid Cloud (combining on-premises data centers with public cloud instances) and Multi-Cloud architectures (using AWS, Azure, and Google Cloud concurrently). Enterprise architects must design robust network topology blueprints, such as dedicated cloud interconnects (AWS Direct Connect, Azure ExpressRoute), unified zero-trust identity federation, and multi-cloud data synchronization.
Technology Standards Catalog & Lifecycle Governance
To prevent technical debt, shadow IT, and vendor lock-in, Phase D establishes a Technology Standards Catalog derived from the TOGAF Technical Reference Model (TRM).
Technology Standardization Framework
- Approved / Standard: Officially supported technologies for all new deployments.
- Emerging / Evaluation: Promising technologies undergoing controlled POC testing.
- Twilight / Phasing Out: Legacy technologies approved for maintenance but barred from new projects.
- Retired / Prohibited: Unauthorized technologies that must be removed immediately due to security or support end-of-life.
Technology Lifecycle Management (N-1 Strategy)
To maintain system stability, architects enforce an N-1 Versioning Strategy, where production environments run strictly on the current stable version ($N$) or immediately preceding minor version ($N-1$). Proactive obsolescence tracking ensures that operating systems, database engines, and hardware components are upgraded prior to vendor End-of-Life (EOL) deadlines.
Network Topology, Security Boundaries & High Availability Patterns
1. Network Topology & Perimeter Security
In Phase D, network architecture defines physical and logical connectivity across compute nodes:
- Demilitarized Zones (DMZs): Multi-tier perimeter zones separating public-facing web tier proxies from internal application servers and database clusters.
- Software-Defined Networking (SDN): Micro-segmentation of virtual network zones, enforcing granular traffic filtering rules between application microservices.
- Zero-Trust Network Architecture (ZTNA): Continuous authentication, mutual TLS (mTLS) encryption, and explicit authorization for every internal inter-service call.
2. Integration with Security Architecture (SABSA Alignment)
TOGAF Phase D integrates closely with SABSA (Sherwood Applied Business Security Architecture):
- Security Services: Encryption in transit (TLS 1.3), encryption at rest (AES-256), Key Management Services (KMS), and Hardware Security Modules (HSMs).
- Identity & Access Management (IAM): Federated single sign-on (SSO), SAML 2.0 / OIDC protocols, role-based access control (RBAC), and privileged access management (PAM).
3. High Availability (HA) & Disaster Recovery (DR) Patterns
Physical and virtual infrastructure must satisfy non-functional Availability (RTO and RPO) requirements:
- Active-Active Multi-Region Clustering: Concurrent workload processing across geographically separated cloud regions, providing zero RTO for critical services.
- Active-Passive / Warm Standby: Secondary passive environment updated continuously via database replication, failing over automatically if primary region suffers outage.
- Recovery Point Objective (RPO) & Recovery Time Objective (RTO): Quantitative targets driving backup cadence, transaction log shipping, and automated failover orchestration.
What is the primary role of Platform Services in TOGAF Technology Architecture?
Under the Cloud Computing PaaS (Platform as a Service) model, which layer remains the responsibility of the customer enterprise architecture team?
Why do enterprise architects establish a Technology Standards Catalog within Phase D?