Cloud service models, OCI physical architecture, and high availability
What this covers
The benefits and use cases of OCI, cloud service models, and how OCI organizes its global infrastructure into regions, availability domains, and fault domains.
Cloud Service Models
IaaS
Infrastructure as a Service — virtualized compute, storage, and networking. The customer manages the OS and applications (e.g., OCI Compute, Block Volume, VCN).
PaaS
Platform as a Service — a managed runtime where you deploy applications without managing the OS (e.g., Autonomous Database, OKE).
SaaS
Software as a Service — fully managed applications consumed directly (e.g., Oracle Fusion Cloud Applications).
OCI Physical Architecture
Construct
Definition
Protects Against
Region
A localized geographic area containing one or more availability domains
Region-wide disasters (use multiple regions for DR)
Availability Domain (AD)
An isolated, fault-tolerant data center within a region with independent power, cooling, and networking
Failure of an entire data center within a region
Fault Domain (FD)
A grouping of hardware within a single AD (three per AD)
Hardware failures within an availability domain
Realm
A logical collection of isolated regions (e.g., commercial, government)
Cross-realm data isolation requirements
Key Distinction
Different fault domains protect against hardware failure inside one AD. Multiple availability domains protect against an AD outage within a region. Different regions are required to survive a region-wide disaster (disaster recovery).
Shared Responsibility Model
Oracle secures the cloud infrastructure: physical data centers, hardware, and the virtualization/hypervisor layer
The customer secures their own data, identity and access configuration, and how they configure OCI services
Security responsibility is divided by layer — it is not equally shared, and the customer never manages Oracle's physical facilities
Benefits of OCI Cloud
Elasticity: scale resources up during peak demand and back down afterward to match actual usage
Pay-as-you-go: shift spending from upfront capital expenditure (CapEx) to operational expenditure (OpEx)
Faster time to market: provision resources on demand in minutes instead of weeks of hardware procurement
Consistent pricing: the same resource costs the same across all regions, providing cost predictability
Managed services: Oracle offloads hardware maintenance, patching, and facility management
