Over the last few years, Oracle GoldenGate has quietly transformed from a traditional replication tool into a powerful, cloud-ready, AI-aware, and streaming-centric data movement platform. If the last time you looked at GoldenGate was during the 12c era, you’re in for a surprise.
This blog walks through the evolution of GoldenGate—from its architectural shift to microservices, through cloud and heterogeneous replication, all the way to the massive feature set introduced in GoldenGate 23ai. We’ll keep it practical, human, and focused on what actually matters in real production environments.
A Quick Look Back: Why GoldenGate Had to Evolve
The last major Oracle University GoldenGate course before recent years was built around GoldenGate 12c. At that time, GoldenGate was already trusted for zero-downtime migrations and real-time replication, but the data landscape itself has changed dramatically since then.
Organizations now expect:
- Cloud-native deployments
- Cross-platform and cross-cloud replication
- Real-time analytics and streaming
- Security, observability, and automation by default
GoldenGate’s evolution directly mirrors these expectations.
The Big Shift: Microservices Architecture (GoldenGate 12.3)
One of the most important milestones was GoldenGate 12.3, which introduced the microservices architecture.
Instead of managing replication with command-line tools and monolithic processes, microservices brought:
- A modern Web UI
- REST APIs for automation
- Independent, scalable services
- Easier lifecycle management
This architectural change laid the foundation for everything that followed—cloud integration, observability, streaming, and AI-driven features.
Cloud-Ready Replication: GoldenGate 18c and 19c
Native Support for Autonomous Databases
With GoldenGate 18c, Oracle introduced native support for Autonomous Transaction Processing (ATP) and Autonomous Data Warehouse (ADW). This meant GoldenGate could now integrate seamlessly with Oracle’s self-managing cloud databases—no hacks, no workarounds.
Cross-Endian Remote Capture (GoldenGate 19c)
GoldenGate 19c removed a long-standing barrier: cross-endian capture.
This made it dramatically easier to capture data from legacy platforms like:
- Solaris
- AIX
- HP-UX
…and replicate directly into the cloud. For organizations modernizing from older UNIX systems, this was a game changer.
Unified Build: Simpler Installs, Fewer Headaches
Another underrated but hugely impactful change in 19c was the unified build.
Before this, GoldenGate installs were tightly coupled to:
- Database versions
- Platform specifics
Now, you install GoldenGate once—no worrying about exact database versions. It simplified upgrades, reduced errors, and saved countless hours for DBAs.
GoldenGate 23ai: A Massive Leap Forward
GoldenGate 23ai is not a “small incremental release.” It’s a platform-level upgrade touching AI, performance, security, streaming, and observability.
Let’s break down the highlights.
AI, Vector Data, and Heterogeneous Streaming
GoldenGate 23ai introduces support for database vector types, opening the door for AI-driven workloads.
What’s powerful here is heterogeneity:
- Oracle → Oracle vectors
- PostgreSQL → Oracle vectors
- Oracle → non-Oracle targets
Just like traditional GoldenGate replication, vector data isn’t locked into a single ecosystem.
A Faster, Smarter Web UI
In large enterprises, GoldenGate environments often run dozens or hundreds of Extracts and Replicats. Older UIs struggled to keep up.
The redesigned UI in 23ai is:
- Optimized for high-process environments
- Faster to refresh
- Easier to troubleshoot at scale
This matters when you’re managing mission-critical replication with tight SLAs.
Calling Database Functions from GoldenGate
Previously, GoldenGate offered around 50–60 built-in transformation functions. Useful—but limited.
Now, GoldenGate can:
- Call database functions
- Execute stored procedures
- Use package functions
- Return values for real-time transformation
This bridges the gap between replication and business logic, without pushing everything into downstream applications.
Enterprise-Grade Security and Identity Integration
GoldenGate now integrates with modern identity providers:
- Token-based authentication
- Azure Active Directory
- Enterprise SSO platforms
This is critical for regulated environments where replication tools must follow the same security standards as the rest of the stack.
Oracle Database 23ai + GoldenGate: Lock-Free Reservation Columns
One of the most exciting Oracle 23ai features is lock-free reservation columns.
In simple terms:
- Multiple transactions can update the same row at the same time
- No traditional row-level locking
- Ideal for high-concurrency workloads
Think shopping carts, inventory systems, or real-time counters.
⚠️ Important note:
This feature is supported only for Oracle-to-Oracle replication, because both source and target must understand the same lock-free data semantics.
Better Observability: Know Exactly What GoldenGate Is Doing
GoldenGate diagnosability took a huge step forward:
- AWR reports now include 7 GoldenGate-specific sections
- Deep insight into process behavior inside the database
- New Replication Performance Advisor
Best of all, this advisor is integrated directly into the Web UI, eliminating the need to manually run database procedures.
Per-PDB Extracts: Better Security and Performance
In multitenant databases, redo data used to be shared across all PDBs. Now, GoldenGate supports per-PDB Extracts.
Benefits:
- Improved security isolation
- Better performance
- Cleaner configurations
Each Extract can focus on a single pluggable database—exactly what modern containerized databases need.
Non-Oracle Enhancements: PostgreSQL and Beyond
GoldenGate isn’t just about Oracle.
For PostgreSQL, 23ai introduces:
- Precise instantiation using Postgres snapshots
- Reduced collision handling
- Cleaner initial loads
This significantly improves Postgres-to-Postgres replication reliability.
From Big Data to Distributed Applications & Analytics
GoldenGate for Big Data has been renamed to GoldenGate Distributed Applications and Analytics (DAA)—a much more accurate name.
New capabilities include:
- Databricks integration
- Google Pub/Sub delivery
- Redis and BigQuery targets
- Stage-and-merge for BigQuery
- XAG support for high availability
If a node fails, GoldenGate restarts automatically on another cluster node—following Oracle MAA best practices.
Streaming Data and AsyncAPI: A New Way to Consume Changes
One of the most innovative additions is AsyncAPI and CloudEvent streaming.
You can now:
- Subscribe to GoldenGate trail files
- Filter by tables and conditions
- Publish changes to messaging platforms
- Build event-driven architectures
This turns GoldenGate into a real-time data publisher, not just a replication engine.
Supported Platforms: The Scope Is Massive
GoldenGate supports:
- Oracle 23ai and earlier (all editions)
- Autonomous Databases
- Exadata, ODA, OCI Base DB
- Amazon RDS, Aurora, Redshift, S3
- Azure Cosmos DB, Data Lake, MySQL, Postgres
- Snowflake, Apache Cassandra, MariaDB
- Streaming platforms, NoSQL stores, SaaS apps
To see the full list, search for “GoldenGate Certification Matrix”—it’s extensive and regularly updated.
The GoldenGate Product Family
GoldenGate is not one product—it’s an ecosystem:
- GoldenGate for Oracle & non-Oracle databases
- GoldenGate for Mainframe
- OCI GoldenGate (now expanding to Azure & Google Cloud)
- GoldenGate Free (perfect for developers)
- GoldenGate DAA
- GoldenGate Veridata (data comparison and repair)
- GoldenGate Stream Analytics
- GoldenGate Migration Utility
Together, they cover replication, streaming, analytics, validation, and modernization.
Final Thoughts
GoldenGate has evolved into a data movement platform for the AI, cloud, and streaming era. Whether you’re modernizing legacy systems, building real-time analytics pipelines, or supporting multi-cloud strategies, GoldenGate 23ai delivers flexibility at an enterprise scale.
If Lesson 1 was about understanding what’s possible, the real value of GoldenGate is how much of that power you can unlock with the right architecture—and the confidence that it’ll keep up with what’s next.
- What Is Oracle GoldenGate? A Beginner-Friendly Introduction
- Oracle GoldenGate 19c Microservices Architecture (MA) – Software Installation Guide
- Oracle GoldenGate 19c Microservices Architecture – Creating Certificates
- Oracle GoldenGate 19c Microservices Architecture – Configuring MA Deployment Using oggca.sh
- Oracle GoldenGate 19c MA Distribution Server: A Complete Beginner-Friendly Guide
- Oracle GoldenGate Microservices Architecture: Receiver Server (Complete Guide)
- Oracle GoldenGate Initial Load Configuration: A Complete Guide for DBAs (2026 Edition)
