PROVIDER Β· The enterprise commercial RDBMS benchmark
Oracle Database
Oracle Database remains the enterprise benchmark, with RAC active-active, Data Guard disaster recovery, and PL/SQL in-database logic forming the standard for financial and government core systems at the price of expensive licensing and lock-in.
4.5pts
Pros
Cons
Product Information
Detailed Notes
Overview
Oracle Database shipped its first commercial SQL relational release in 1979 from Larry Ellison's team and is now maintained by Oracle, with 19c and 23ai as the long-term releases, holding a durable lead in core transactional systems for finance, telecom, and government.
For Backend & Database work its strength is extreme stability and completeness, since RAC provides shared-storage active-active, Data Guard delivers cross-site disaster recovery, PL/SQL pushes business logic into the database, and partitioning with parallel execution handles very large data, effectively a prerequisite in regulated high-availability industries.
The cost is price and lock-in, as licensing per CPU core or named user with an EE core factor often of 0.5 pushes annual fees into tens of thousands of dollars upward, proprietary syntax binds applications to the ecosystem so migration requires rewriting stored procedures, and operations demand a dedicated DBA team.
Core Strengths
- RAC provides shared-storage active-active with node failure transparency and scale-out capacity
- Data Guard delivers physical or logical standby with zero-data-loss cross-site recovery
- PL/SQL and built-in packages keep business logic and batch processing inside the database
- Partitioning, parallel execution, materialised views, and AWR diagnostics handle very large workloads
Product Ecosystem
Database 19c and 23ai
19c has the broadest coverage as long-term support while 23ai adds AI Vector Search, and the certified version of your application vendor should drive the choice.
RAC and Data Guard
RAC addresses same-city active-active and compute scaling while Data Guard covers remote disaster recovery, typically combined for core systems.
Exadata and Autonomous Database
Exadata is the engineered system combining hardware and software, while Autonomous Database is the fully managed service on OCI that cuts routine operations sharply.
Drawbacks
- Licensing is expensive, per-core with enterprise core factors and options making totals hard to predict
- Proprietary syntax and PL/SQL create strong lock-in, so migration means rewriting and retesting every procedure
- A dedicated DBA team is required, with higher cost and hiring difficulty than open source
- Cloud cross-region and egress costs run high, complicating hybrid architectures
Use Cases
- Core transaction systems in finance and government (β β β β β ) - Mature RAC and disaster recovery
- Strict compliance and auditing (β β β β β ) - Complete audit and security options
- Very large data warehouses (β β β β ) - Strong partitioning and parallelism
- Budget-sensitive small projects (β ) - Licensing is unaffordable
Pricing
| Plan | Price | Scope |
|---|---|---|
| Express edition | $0 | Development and limited commercial use with low caps |
| Standard Edition 2 | Per-core annual licence | Up to 16 threads without RAC |
| Enterprise Edition | Per-core licence plus options | RAC, partitioning, advanced compression |
Oracle cost is almost entirely licensing, with Enterprise Edition billed per core plus separate options for RAC, partitioning, and advanced compression, commonly reaching tens of thousands of dollars a year before roughly 22 percent support fees, so budget licence, support, DBA time, and infrastructure together.
FAQ
-
When is Oracle mandatory? Mainly when the application vendor only certifies Oracle, RAC-level active-active is required, or strict compliance auditing applies, since greenfield work usually has cheaper options.
-
Can it migrate to PostgreSQL? Yes, with effort in rewriting PL/SQL, tuning around data type and optimiser differences, and regression testing, so start by migrating a read-only reporting replica to validate.
-
Is Autonomous Database economical? Operations drop clearly and it suits teams without a DBA, while per-OCPU and storage billing needs a detailed comparison against self-built RAC for sustained heavy load.