Architecture models

This objective covers architectural models and how structural design decisions affect an organization’s security posture, cost, and resilience. CompTIA tests your ability to weigh the security trade-offs of different infrastructures, ranging from hyper-modern cloud systems to isolated industrial environments.

1. Cloud vs. On-Premises Architecture

The fundamental choice of where your data and computing power physically live dictates your operational risk profile.

On-Premises vs. Cloud

• On-Premises: The organization owns and maintains the physical data center, hardware, and networking gear. You have total architectural control, but you bear all the operational costs, power management, physical security, and recovery burdens.
• Cloud: Infrastructure is hosted by a third-party provider (like AWS, Azure, or GCP). You gain massive scalability and ease of deployment, but you introduce risks concerning third-party vendors and data visibility.

The Cloud Responsibility Matrix (Shared Responsibility Model)

When moving to the cloud, security tasks are split between the provider and the customer based on the service model. A common exam pitfall is assuming the cloud provider handles everything.

• Infrastructure as Service (IaaS): The provider secures the physical facilities, bare-metal servers, and hypervisor. You are responsible for the operating system, patching, firewalls, applications, and data.
• Platform as a Service (PaaS): The provider secures the physical infrastructure and manages the OS and runtime environment (e.g., AWS Elastic Beanstalk or Azure SQL). You only manage your application code and data.
• Software as a Service (SaaS): The provider manages the entire stack from the physical infrastructure up to the application interface (e.g., Microsoft 365). Your primary responsibility is data governance, user identities, and access control permissions.

Hybrid Considerations

A hybrid model blends on-premises infrastructure with public cloud services. While highly flexible, it introduces major security implications, including complex identity synchronization (bridging local Active Directory and cloud IAM) and the risk of misconfiguring the secure network tunnels (such as IPsec VPNs or dedicated circuits) that connect the two environments.

2. Modern Abstraction: Microservices, Containers, & Serverless

Modern application architecture breaks monolithic code into highly distributed, decoupled components.

• Microservices: Breaking a single application into dozens of small, independent services that communicate via lightweight APIs.
  • Security Implication: If a single service is compromised, the attacker doesn’t automatically gain control of the entire application. However, you dramatically expand the network attack surface because you must secure all inter-service API communications.
• Containerization: Bundling an application and all its dependencies into an isolated package (e.g., a Docker container) that shares the host OS kernel.
  • Security Implication: Containers are incredibly lightweight and enforce scalability. However, because they share the same underlying OS kernel, a kernel-level vulnerability on the host can lead to container breakout attacks.
• Serverless (Function-as-a-Service – FaaS): Developers deploy individual snippets of code that run in response to specific events (e.g., AWS Lambda). The cloud provider spins up a micro-container, executes the function, and instantly tears it down.
  • Security Implication: Traditional host-based tools (such as standard endpoint protection or local firewalls) cannot be installed because you have no access to the underlying OS. Security must shift entirely to strict API gateway routing, robust input validation, and highly restricted IAM permissions.
• Infrastructure as Code (IaC): Using machine-readable configuration files (like Terraform, Ansible, or CloudFormation templates) to automatically provision and manage infrastructure.
  • Security Implication: Drastically reduces manual misconfigurations. Security teams can run automated security scanning tools directly against the IaC templates before deployment, catching open ports or weak access rules before the infrastructure is even built.

3. Network Infrastructure & Segment Isolation

How network boundaries are drawn impacts how easily an attacker can pivot through an environment.

• Physical Isolation (Air-Gapped): A network security measure in which a computer system or network is completely isolated from any other network, including the public internet.
  • Security Implications: Provides maximum protection against remote cyberattacks. The only way to compromise or extract data from an air-gapped system is through physical proximity or a direct human vector (such as an unauthorized USB device).
• Logical Segmentation: Using virtual technologies (like VLANs or software ACLs) to isolate systems that share the same physical cables and switches.
• Software-Defined Networking (SDN): Completely decoupling the network routing/forwarding control plane from the physical hardware switches and moving it into centralized management software.
  • Security Implication: Allows for highly dynamic microsegmentation and automated firewall rule updates across thousands of endpoints instantly. However, if an attacker compromises the central SDN controller, they gain absolute control over the entire enterprise network topology.

4. Centralized vs. Decentralized Models

• Centralized Architecture: All processing, data storage, and security controls are managed from a core, singular location (e.g., a central mainframe or a single primary cloud tenant).
  • Trade-off: Highly manageable and consistent configuration enforcement, but it creates a single point of failure.
• Decentralized Architecture: Distributing processing and workloads across multiple independent nodes or geographic locations (e.g., Edge computing, branch offices).
  • Trade-off: High resilience and performance responsiveness, but vastly more difficult to monitor, patch, and maintain a consistent security baseline across all nodes.

5. Specialized & Operational Technology (OT) Systems

Unlike standard IT environments that prioritize Confidentiality, Operational Technology environments prioritize Availability and Safety.

• Internet of Things (IoT): Everyday physical devices embedded with internet connectivity (smart thermostats, connected security cameras).
  • Security Implications: These devices frequently suffer from an inability to patch, short lifecycles, and weak default configurations, turning them into prime targets for botnet recruitment.
• Industrial Control Systems (ICS) / SCADA: Large-scale distribution networks for monitoring and controlling physical industrial infrastructure, such as power grids, water treatment plants, and assembly lines.
  • Security Implication: These networks rely heavily on legacy applications and protocols that lack basic authentication or encryption. A compromise here can lead to catastrophic real-world physical damage.
• Real-Time Operating System (RTOS): An operating system designed to process data and events with absolutely zero buffering delay or jitter. It is used when precise execution timing is mandatory.
  • Examples: Vehicle anti-lock braking software, pacemakers, and military missile guidance systems.
• Embedded Systems: A specialized microcomputer system designed to perform a single, dedicated task, often embedded within a larger mechanical device (such as a smart appliance, engine control module, or network interface card).

6. Architecture Considerations & Trade-offs

When designing an architecture model, architects must constantly balance these competing design pillars:

Consideration	Security & Operational Reality
Availability / Resilience	High availability requires building redundant components. If a primary system fails, a secondary system takes over automatically (ease of recovery), ensuring zero business disruption.
Cost	Building deep defense-in-depth layers, maintaining multi-region cloud backups, and investing in dedicated hardware security tools adds significant financial overhead.
Responsiveness	Moving security inspection tools directly into the live data path (such as deep packet inspection) can introduce millisecond-level latency, slowing system performance.
Scalability	Cloud, container, and serverless architectures excel at dynamically scaling to meet demand, but rapidly expanding systems can quickly outpace visibility if logging tools aren’t configured to scale with them.
Risk Transference	Choosing a SaaS or cloud vendor transfers specific infrastructure security risks to the provider, though the organization ultimately remains accountable for its own data.
Patch Availability vs. Inability to Patch	Standard IT systems receive weekly automated software updates. Specialized OT, embedded, and IoT systems often have an absolute inability to patch due to vendor abandonment or because the system cannot afford the downtime required for a reboot.
Power & Compute	Edge nodes, mobile devices, and embedded systems have strict physical engineering constraints. They cannot run heavy local security agents or process complex cryptographic calculations because they lack the necessary compute power and power (battery) resources.