Study guide
6.2 OCI Base Database & Exadata Database Services
Oracle Cloud Infrastructure Foundations Associate 1Z0-1085-26
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: Getting Started with OCI & Cloud Architecture
1.1 Cloud Computing Principles & Deployment Models 1.2 OCI Global Architecture: Regions, Availability Domains & Fault Domains 1.3 OCI Tenancy & Compartment Hierarchy 1.4 OCI Distributed Cloud Strategy & Multicloud Offerings
2 Chapter 2: OCI Compute & Serverless Services
2.1 Compute Foundations: Bare Metal, Dedicated VM Hosts & Virtual Machines 2.2 Compute Shapes, Images & Autoscaling 2.3 Containers & Serverless: OKE, Container Instances & Oracle Functions
3 Chapter 3: OCI Virtual Cloud Network & Core Connectivity
3.1 Virtual Cloud Network (VCN) Architecture & Subnets 3.2 OCI Network Gateways & Hybrid Connectivity
4 Chapter 4: OCI Network Security, Routing & Load Balancing
4.1 Network Security Controls, Route Tables & FastConnect 4.2 OCI Load Balancing Services
5 Chapter 5: OCI Storage Services
5.1 OCI Block Volume Service 5.2 OCI Object Storage Service 5.3 OCI File Storage Service (FSS)
6 Chapter 6: OCI Database Solutions
6.1 OCI Autonomous Database 6.2 OCI Base Database & Exadata Database Services 6.3 OCI Managed MySQL & NoSQL Database Services
7 Chapter 7: OCI Identity and Access Management (IAM)
7.1 OCI Shared Security Responsibility Model 7.2 OCI IAM Identity Domains & Identity Constructs 7.3 OCI IAM Policy Syntax, Inheritance & Structure
8 Chapter 8: OCI Platform Security & Protection
8.1 OCI Vault, Key Management & Data Encryption 8.2 OCI Cloud Guard & Security Zones 8.3 OCI Perimeter Protection, Vulnerability Scanning & Bastion Service
9 Chapter 9: OCI Pricing, Cost Management & Governance
9.1 OCI Pricing Models & Free Tier 9.2 OCI Cost Management & Budgeting Tools 9.3 Resource Governance, Service Limits & Financially Backed SLAs
6.2 OCI Base Database & Exadata Database Services
Key Takeaways Explain This Topic AI
Co-managed database services in OCI provide full administrative control, root operating system access, and custom software configuration capabilities while leveraging Oracle cloud automation for lifecycle tasks.
Base Database Service offers Virtual Machine DB Systems (supporting 1-node or 2-node Real Application Clusters on remote block storage) and Bare Metal DB Systems (single-tenant physical hardware with direct-attached NVMe storage).
Oracle Real Application Clusters (RAC) provides active-active multi-node clustering over shared storage, delivering high availability, fault tolerance, and workload balancing across instances with zero downtime.
Exadata Database Service delivers extreme performance on dedicated engineered systems utilizing intelligent storage servers, 100 Gbps RoCE fabrics, Smart Scan SQL offloading, and Hybrid Columnar Compression (HCC).
Customers choose between License Included and Bring Your Own License (BYOL) commercial models, and select from Standard, Enterprise, High Performance, and Extreme Performance software editions.
Last updated: September 2026
6.2 OCI Base Database & Exadata Database Services
Note
Foundations Blueprint Focus : While Autonomous Database handles fully managed PaaS workloads, many enterprise applications require co-managed databases where the customer retains root operating system access and administrative control. The OCI Foundations Associate (1Z0-1085-26) exam tests your ability to distinguish between Base Database Service and Exadata Database Service, understand Oracle Real Application Clusters (RAC), and evaluate licensing models (License Included vs. BYOL).
Not all enterprise applications can or should run on a fully managed autonomous platform. Many commercial off-the-shelf enterprise packages, legacy business systems, and specialized database architectures require custom operating system configurations, specific third-party monitoring agents, or particular legacy database versions. For these scenarios, Oracle Cloud Infrastructure provides co-managed Oracle Database services .
The Co-Managed Paradigm: Freedom with Cloud Automation
In a co-managed service model, Oracle and the customer establish a shared administrative boundary:
Oracle's Responsibility : Oracle provides the physical cloud infrastructure, hypervisors, physical networking, and cloud automation tooling. Through the OCI Console, CLI, and REST APIs, Oracle provides one-click lifecycle automation for database provisioning, on-demand CPU scaling, automated backup scheduling, software patching, and disaster recovery configuration.
Customer's Responsibility : Unlike Autonomous Database, the customer receives full administrative access , including root operating system credentials ( opc and root access via SSH) and complete database administrator ( SYSDBA ) privileges.
When to Choose Co-Managed Services
Organizations select co-managed database services when their technical or operational requirements include:
Root Operating System Access : The application requires custom OS kernel tuning, local shell scripts, cron jobs, or specific filesystem mounting configurations.
Third-Party Agent Installation : Security policies require installing host-based intrusion detection agents, custom APM monitoring probes (e.g., Datadog, Dynatrace), or enterprise backup agents directly on the database host.
Legacy Database Versions : The enterprise must run specific database versions (such as Oracle Database 11g, 12c, 19c, or 21c) to maintain vendor certification for legacy commercial software.
Custom Database Features & Initialization : The database relies on non-standard initialization parameters, custom database options, or direct access to internal database trace files.
Base Database Service
OCI Base Database Service allows organizations to run genuine Oracle Database Enterprise Edition or Standard Edition on standard OCI compute shapes. Base Database Service offers two distinct infrastructure deployment architectures: +-------------------------------------------------------------------------+
- | BASE DATABASE SERVICE OPTIONS |
- +------------------------------------+------------------------------------+
- | VIRTUAL MACHINE DB SYSTEMS | BARE METAL DB SYSTEMS |
- +------------------------------------+------------------------------------+
- | • Multi-tenant or dedicated VMs | • Single-tenant physical hardware |
- | • Remote OCI Block Volumes | • Locally attached NVMe flash |
- | • 1-Node or 2-Node RAC | • Single-node architecture |
- | • Dynamic CPU & storage expansion | • Extreme raw I/O throughput |
- | • Cost-effective general workloads | • Dynamic CPU core allocation |
- +------------------------------------+------------------------------------+
- 1. Virtual Machine (VM) DB Systems
Virtual Machine DB Systems run on OCI VM compute shapes and use network-attached OCI Block Volumes for database storage.
Single-Node DB Systems : Runs a single database instance on a single VM. Highly cost-effective for development, testing, staging, and non-mission-critical production environments.
2-Node Real Application Clusters (RAC) : Deploys two synchronized VM compute instances across distinct Fault Domains (or across Availability Domains in multi-AD regions) connected to shared block storage, delivering active-active enterprise clustering.
Elastic Scaling : Compute cores can be scaled dynamically without rebuilding storage, and block storage capacity can be expanded online up to 80 TB without service interruption.
2. Bare Metal (BM) DB Systems
Bare Metal DB Systems run directly on dedicated, single-tenant physical servers without any virtualization hypervisor overhead.
Extreme Storage Performance : Equipped with locally attached, high-speed NVMe solid-state drives , delivering raw I/O performance with sub-millisecond latencies for I/O-intensive enterprise workloads.
Dynamic Core Sizing : Provides dynamic CPU scaling through software allocation. A customer can provision a 52-core Bare Metal server but enable and pay software licenses for only a subset of cores (e.g., 16 cores), expanding active cores as database transaction volume grows.
Single-Node Architecture : Bare Metal DB Systems support single-node deployments (RAC is not available on Bare Metal shapes; RAC requires VM DB Systems or Exadata).
Oracle Real Application Clusters (RAC)
Oracle Real Application Clusters (RAC) is Oracle's premier active-active database clustering technology. In a traditional database setup, a single database instance runs on a single server. If that server fails, the database is unavailable until a standby server boots up or restarts the database.
In an Oracle RAC environment:
Active-Active Clustering : Multiple independent server nodes concurrently access and modify a single, shared database located on shared storage (managed by Oracle Automatic Storage Management / ASM).
Cache Fusion Technology : RAC utilizes a dedicated, ultra-low-latency private interconnect network between cluster nodes. Through Cache Fusion, memory blocks are synchronized directly between the nodes' system memory (SGA) over the interconnect, eliminating the need to write blocks to disk to share data between instances.
High Availability & Fault Tolerance : If one cluster node experiences a physical hardware crash, operating system panic, or network disruption, surviving nodes continue processing transactions seamlessly. Application client sessions automatically failover to surviving nodes via Transparent Application Failover (TAF) with zero data loss and uninterrupted service.
Zero-Downtime Maintenance : DBAs can perform rolling operating system upgrades, security patches, and hardware maintenance by taking down one node at a time while the remaining node carries production traffic.
Important
In OCI Base Database Service, Oracle RAC is supported exclusively on 2-Node Virtual Machine DB Systems . It requires the Enterprise Edition Extreme Performance software edition.
Exadata Database Service
Exadata Database Service represents the pinnacle of enterprise database architecture in Oracle Cloud Infrastructure. Exadata is not commodity hardware running database software; it is a purpose-built, co-engineered hardware and software platform optimized specifically for Oracle Database workloads. +-------------------------------------------------------------------------+
- | EXADATA DATABASE SERVICE ARCHITECTURE |
- | +-------------------------------------------------------------------+ |
- | | Dedicated Database Compute Nodes (Run DB Instances & Oracle RAC) | |
- | +----------------------------------+--------------------------------+ |
- | | 100 Gbps RoCE Network Fabric
- | v
- | +-------------------------------------------------------------------+ |
- | | Intelligent Exadata Storage Servers (Smart Scan Predicate Offload)| |
- | | • Exadata Smart Flash Cache • Hybrid Columnar Compression | |
- | +-------------------------------------------------------------------+ |
- +-------------------------------------------------------------------------+
- Architectural Composition
Exadata separates database processing from storage processing using dedicated, specialized tiers:
Database Compute Nodes : High-capacity enterprise servers running database compute instances, query coordination, and Oracle RAC.
Intelligent Storage Servers : Specialized storage engines equipped with dedicated CPUs, high-speed NVMe flash memory, and intelligent Exadata Storage Server software.
100 Gbps RoCE Network Fabric : Compute nodes and storage servers are interconnected using RDMA over Converged Ethernet (RoCE) switches delivering 100 Gbps bandwidth and sub-2-microsecond network latencies with zero host CPU overhead.
Proprietary Exadata Software Innovations
Exadata achieves orders-of-magnitude faster performance than generic cloud databases due to three proprietary technologies:
Smart Scan (Cell Offloading) : In traditional cloud databases, when a query executes (e.g., SELECT customer_name FROM orders WHERE order_total > 500 ), the entire orders table must be copied over the network from storage to database compute servers, where filtering occurs. With Exadata Smart Scan, the SQL query predicate ( WHERE order_total > 500 ) and column projection are offloaded directly to the intelligent storage servers. The storage processors filter the data in hardware and return only the matching rows and columns to the compute nodes, reducing internal network traffic by up to 90%.
Hybrid Columnar Compression (HCC) : Patented compression technology that organizes data into Compression Units within database blocks. It delivers 10x to 15x compression ratios for historical and analytical data, drastically reducing storage footprint while accelerating query speeds by reading compressed data directly from disk.
Exadata Smart Flash Cache : Intelligent software algorithms automatically identify hot, frequently accessed data blocks and cache them in ultra-fast PCIe NVMe flash memory, bypassing slower disk storage entirely.
Deployment Options: Cloud Infrastructure vs. Cloud@Customer
Exadata Database Service is available in two cloud delivery models:
Exadata Cloud Infrastructure : Hosted and operated inside OCI public cloud datacenters. Customers dynamically scale database compute servers and storage servers independently on demand.
Exadata Cloud@Customer : Physical Exadata engineered hardware is delivered and deployed directly inside the customer's on-premises datacenter , but connected to OCI and operated by Oracle as a fully managed cloud service. This model satisfies strict data sovereignty laws, corporate data residency requirements, and microsecond latency needs for on-premises legacy systems while providing cloud pay-as-you-go economics.
Licensing Models & Software Editions
OCI offers flexible commercial models and software tiers to accommodate diverse organizational requirements:
Licensing Models
License Included : The hourly cloud billing rate bundles the underlying cloud infrastructure, Oracle Database software licenses, and 24/7 Oracle Premier Support. Ideal for organizations building new cloud-native applications or expanding database footprints without existing software investments.
Bring Your Own License (BYOL) : Customers repurpose existing on-premises Oracle Database licenses to OCI. BYOL significantly lowers hourly operational costs because customers pay only for the underlying infrastructure and cloud automation services.
Database Software Editions
Base Database Service offers four progressive software editions:
Edition Capabilities & Included Options
Standard Edition Base Oracle Database engine suitable for departmental applications with moderate concurrency. Limited to single-node VM deployments.
Enterprise Edition Full enterprise relational database engine with advanced security, management packs, and clustering capabilities.
Enterprise Edition High Performance Bundles Enterprise Edition with advanced database options: Multitenant, Partitioning, Advanced Security (TDE), Advanced Compression, Real Application Testing, and Management Packs.
Enterprise Edition Extreme Performance Includes all Oracle Database Enterprise Edition features and options: Oracle RAC, In-Memory Database, Active Data Guard, and all management packs. Mandatory for 2-Node RAC VM DB Systems and Exadata .
Architectural Comparison: Base DB vs. Exadata DB Service
- Architectural Dimension Base DB: Virtual Machine Base DB: Bare Metal Exadata Database Service
- Tenancy Multi-tenant or dedicated VM Single-tenant physical host Dedicated engineered Exadata hardware
Storage Architecture Network-attached Block Volumes Locally attached NVMe SSDs Intelligent Exadata Storage Servers
- Oracle RAC Support Yes (2-Node RAC VM) No (Single-node only) Yes (Multi-node active-active RAC)
- Smart Scan Offloading Not supported Not supported Supported (Proprietary cell offload)
- Hybrid Columnar Compression Not supported Not supported Supported (10x-15x compression)
- OS / Root Access Yes (Full root & SYSDBA) Yes (Full root & SYSDBA) Yes (Full root & SYSDBA)
- Maximum Scalability Up to 80 TB block storage Local NVMe storage Petabyte-scale multi-rack clusters
Primary Use Case Cost-effective dev/test & mid-tier High raw I/O single-node apps Mission-critical enterprise OLTP & DW