Oracle GoldenGate 23ai
Welcome to a new blog series about Oracle GoldenGate. For over 20 years Oracle GoldenGate products have helped deliver, validate and secure data. It is a tool for moving and integrating data in real-time.
Oracle GoldenGate enables real-time data capture and facilitates the movement of data across different locations. It is a cloud-agnostic software that can be run on-premises or in the cloud. Oracle GoldenGate can be used to transfer between multiple sources and targets, these sources and targets include a wide range of databases such as Oracle Cloud, Azure Data Lake, Snowflake and more, as well as being able to interact with other data stores (cloud storage, parquet files, etc...).
Let's start by exploring the setup process, key features, and use cases of GoldenGate 23ai. In future blogs, we'll dive into Oracle Veridata 23c, including how to configure jobs for database comparison, as well as running GoldenGate in a Maximum Availability Architecture (MAA) environment.
Setting up a GoldenGate 23ai environment
When setting up GoldenGate 23ai, it's important to consider the environment you're using. The way you set up GoldenGate depends on what kind of sources and targets (databases, parquet files, cloud storage, etc...) you're moving data between. For example, if you're working with Oracle databases, you'll usually use a more classic setup with one service manager and one deployment. However, if you're connecting to other databases, such as MySQL, you'll probably require a different approach.
In our test environment, where we were transferring data between Oracle and MySQL, we first set up a new Service Manager for GoldenGate 23ai. Within this Service Manager, we then created two separate deployments: one for Oracle and another for MySQL. This allowed us to replicate data between two different types of databases with ease. It is also worth noting the importance of preparing your databases beforehand and dedicating time to your pre-requisites. This can be in the form of turning on extra logging on the source database to track changes, patching, creating special GoldenGate user accounts, and making sure you have enough server resources.
For further information on database preparation and service manager/deployment setup, visit the Oracle GoldenGate Documentation.
Features of GoldenGate 23ai
GoldenGate 23ai builds on many powerful features of 21c while introducing new capabilities, including support for AI data types such as vectors.
Here are some of GoldenGate's most prominent features:
- Real-Time Data Replication: GoldenGate captures changes to data as they occur and delivers them to target systems with minimal latency.
- Heterogeneous Database Support: It supports a wide variety of database platforms (Oracle, SQL Server, DB2, MySQL, PostgreSQL, etc.) and operating systems, allowing for data integration across diverse environments.
- Change Data Capture (CDC): GoldenGate efficiently captures database changes (inserts, updates, deletes) without requiring significant overhead on the source system.
- Data Transformation: It provides robust data transformation capabilities, allowing you to modify data during replication to meet the requirements of the target system. This includes filtering, mapping, and data conversion.
- High Availability and Disaster Recovery: GoldenGate can be used to create highly available systems and implement disaster recovery solutions by replicating data to standby databases.
- Zero Downtime Migrations: It facilitates database migrations with minimal to no downtime by replicating data to the new system while the old system is still in operation.
- Security: It offers various security features, such as data encryption and authentication, to protect sensitive data during replication.
As mentioned, GoldenGate 23ai introduces new AI-related features, while also enhancing existing capabilities, such as conflict detection and its introduction of support for newer database releases including MySQL 8.4 and 9.0.
Below is a list of some of GoldenGate 23ai's new features:
- Advanced Streaming Capabilities: GoldenGate 23ai allows for real-time data streaming to popular messaging platforms such as Apache Kafka, Oracle Streaming Service, and Confluent Cloud. This enables event-driven architectures and real-time analytics, opening up new possibilities for data integration.
- Enhanced Conflict Detection and Resolution: This release improves conflict detection and resolution, ensuring data consistency in complex replication scenarios.
- Support for Oracle Database 23ai Features: GoldenGate 23ai leverages new features in Oracle Database 23ai, such as JSON Relational Duality and GoldenGate Data Streams, allowing direct capture of changes on the document side.
- Vector Replication: This feature enhances performance by processing data in vectors, leading to faster replication and reduced latency.
- Direct Connectivity for SQL Server: GoldenGate 23ai adds direct connectivity support for SQL Server, simplifying configuration and improving performance.
To read more about existing and new features of Oracle GoldenGate, visit:
Why use GoldenGate 23ai
If you're already using GoldenGate, upgrading to the latest version offers significant benefits, even without leveraging AI-specific data types. You'll still benefit from significant performance gains via features such as vector replication and direct connectivity enhancements, leading to faster data movement. The advanced streaming capabilities remain valuable for real-time data integration with various platforms, enabling modern data architectures. Furthermore, the improved conflict resolution and streamlined user interface simplify management and enhance data consistency across your systems, regardless of whether or not you're using AI data types.
Here are a few more reasons for using GoldenGate 23ai:
- Using OCI GoldenGate maintenance management you are able to schedule upgrades for your deployments. This is done by defining a start time for the upgrade, if a start time isn't defined the upgrade is scheduled for the weekend closest to the calculated end of the auto upgrade period. As mentioned, the deployment will automatically upgrade if it isn't updated within a certain timeframe, for example, security fix patches have a given timeframe of 14 days.
- There are some limitations with other replication tools that can be circumvented using GoldenGate. For example, AWS DMS has a limitation whereby it no longer supports object names larger than 30 bytes, we’ve recently helped a customer overcome this via the assisted installation of GoldenGate 23ai.
- A key future consideration is leveraging GoldenGate 23ai to keep Retrieval Augmented Generation (RAG) and AI datasets up to date by seamlessly integrating multiple data sources.
To find out more on any of the points discussed within this blog, please use the references below:
Thank you for reading, we hope you found this blog insightful.
For support with your upgrade, contact us at [email protected], and chat with one of our specialists today regarding our GoldenGate services.