23ai New Features
Oracle Database 23ai, the latest long-term support release, introduces over 300 new features designed to enhance AI integration, developer productivity, and data security. This release marks a significant evolution from its predecessor, Oracle 19c, by embedding AI capabilities directly into the database, thereby streamlining application development and data management.
AI Vector Search: Semantic Intelligence Within the Database
A standout feature of Oracle 23ai is AI Vector Search, which enables the storage and querying of vector embeddings representing the semantic content of text, images, and other unstructured data. This functionality facilitates similarity searches and supports Retrieval-Augmented Generation (RAG), allowing seamless integration with large language models (LLMs) such as OpenAI’s GPT-4, Cohere, and Llama2. Developers can utilize over 90 tested embedding models, and the feature is compatible with frameworks like LangChain and LlamaIndex .
JSON Relational Duality Views: Bridging JSON and Relational Models
Oracle 23ai introduces JSON Relational Duality Views, a feature that unifies relational and document data models. This allows developers to interact with the same underlying data as either JSON documents using standard document APIs or as relational data using SQL. This duality simplifies application development by providing flexibility in data representation and manipulation .
SQL Firewall: Enhanced Security for SQL Operations
The new SQL Firewall feature in Oracle 23ai offers robust security by detecting, blocking, and logging unexpected SQL statements and session contexts. This proactive approach helps prevent unauthorized access and potential SQL injection attacks, thereby strengthening the database’s security posture .
True Cache: High-Performance In-Memory Caching
True Cache is an in-memory, consistent, and automatically managed cache for Oracle Database. It operates similarly to an Oracle Active Data Guard reader farm but is designed for performance and scalability. True Cache instances are mostly diskless and can be directly connected to by applications for read-only workloads, enhancing performance and scalability .
Raft Replication: Built-In High Availability
Oracle 23ai incorporates Raft replication, providing native, built-in replication for Oracle Globally Distributed Database. This feature eliminates the need for additional configurations like Oracle GoldenGate or Data Guard. Raft replication uses a consensus-based protocol to ensure high availability and consistency, enabling declarative replication configuration with sub-second failover for seamless continuity .
Developer-Centric Enhancements
Oracle 23ai includes several features aimed at improving developer productivity:
- SQL Analysis Report: Assists developers in writing efficient SQL statements by reporting common issues that can lead to poor performance.
- Schema Annotations: Allows developers to add descriptive metadata to database objects, facilitating better understanding and maintenance.
- Application Usage Domains: Enables the definition of column usage, such as emails or credit cards, enhancing data validation and integrity.
- Boolean Data Type: Introduces a native Boolean data type, simplifying logical operations in SQL.
- Enhanced SQL Syntax: Includes support for aliases in GROUP BY and HAVING clauses, and joins in DELETE and UPDATE statements, improving query flexibility .
In-Memory Enhancements: Optimized Performance
Oracle 23ai enhances its in-memory capabilities with features like Automatic In-Memory (AIM), which can automatically create Join Groups and enable In-Memory Optimized Arithmetic. These enhancements aim to improve query performance and resource utilization .
Comments
Post a Comment