The Challenges of Today's Databases

These are some of the key challenges users experience with managing data in modern IT environments.

New Workloads: Users are trying to manage the influx of new workloads, which often involve new data types. The challenge is determining how to extract business insights when dealing with unstructured data and combining it with graph or vector data.

Security: Security often isn't a priority until a threat occurs. To protect your data from theft and attacks, it's important to take proactive measures. How do you prevent SQL injection? What steps can you take to stop bad actors from compromising your data?

Performance: Users expect their applications to perform efficiently. It is unacceptable to attribute performance issues to the database as the bottleneck.

Complexity: Adding more workloads and incorporating specialized databases into your architecture continuously increases its complexity.

Oracle Database 23ai

Oracle Database 23ai is not just an ordinary database; it is a robust system capable of handling various data types, including JSON, vector, and graph formats. It enables developers to work with unparalleled consistency, scalability, security, and availability. Oracle Database 23ai is designed to simplify data management tasks and enhance data capabilities. Oracle has integrated autonomous operations throughout the database, using machine learning and automation to simplify previously complex configuration and management processes. The Oracle Autonomous Database provides the most efficient way to leverage Oracle's converged database engine.

Workload Support

Oracle Database 23ai supports DBAs in implementing and managing new workloads by providing advanced features that streamline the process.

JSON Relational Duality

The JSON-relational duality in Oracle Database 23ai is a transformative feature, offering developers flexibility and ease of use. This innovation addresses the long-standing challenge of choosing between relational and document models for application development. With this capability, there's no need to search for specialized document databases. You can now combine document and relational data within a single database environment. This advancement improves collaboration between developers and business teams, streamlining workflows within the same database platform.

Data Use Case Domains

This functionality in Oracle Database 23ai introduces a powerful tool for application development. Data use case domains allow you to define how columns are used through distinct data dictionary objects. This approach simplifies the application of domain-specific knowledge across your projects.

Microservice Support

The microservices design approach has gained popularity in recent years, and Oracle Database 23ai has integrated support for automatic compensating transactions. This feature is fundamental to the microservices saga design pattern, simplifying the creation of applications with transactions that span multiple microservices. It ensures these transactions can be securely and efficiently rolled back when necessary. It has also improved the performance of transactional event queues, which are crucial for running microservice applications within the database, resulting in smoother and faster operations. Oracle Database 23ai also supports JavaScript development in stored procedures and has improved its machine learning algorithms and over 400 new features.

AI Vector Search

Oracle Database 23ai introduces a new data type: vector data. There is no AI without data, and supporting vector data in Oracle Database 23ai is essential. Organizations face challenges in integrating enterprise data with large language models, and managing this additional data type can further increase complexity.

Vector Integration: Oracle Database 23ai supports similarity or semantic searches, and the advantage is that you do not need to add another database. The vector is included as an additional data type within Oracle Database 23ai. Managing and moving data can be costly, and incorporating specialized databases adds complexity and technical debt. With Oracle Database 23ai, you can store vector embeddings directly within the Oracle Database and perform similarity searches where your data resides, eliminating the need for data movement. The real strength of the vector data type in the 23ai release is its ability to perform these similarity searches.

Advanced Search Capabilities: In the context of today's AI applications, similarity searches are crucial for quickly locating and matching data points within large datasets. Whether you are working on image recognition, language processing, or recommendation systems, these searches enable accurate, data-driven decisions. Oracle Database 23ai improves this capability by supporting vector search, allowing you to perform similarity searches while leveraging relational predicates. With Oracle Database 23ai, you can conduct similarity searches, query unstructured JSON data, map locations, and analyze relational business data. Also, you can even perform similarity searches on documents stored in Excel.

Scalable Performance: Adding another vector database to your infrastructure is possible but comes with challenges. Joining queries between two different systems requires additional integration work. This may lead to the need for more staff to manage the complexity and ensure data remains synchronized across both systems. Regarding performance, Oracle has decades of experience with scalable, high-performance technologies such as RAC, Partitioning, and Exadata. By using Oracle Database 23ai, you can leverage these advanced features and core database capabilities for your AI workloads.

Security

Oracle Database 23ai introduces major enhancements that strengthen database security and simplify developer workflows. Here is an overview of the key security features:

Developer Role

Oracle Database 23ai introduces a dedicated role specifically for application developers. This role is designed to provide all the necessary privileges for designing, implementing, and deploying applications. This approach reduces security risks by removing the need for DBAs to grant developers broad, potentially unnecessary permissions.

Schema Privileges

Oracle Database 23ai allows system privileges to be granted at the schema level. This means admins can apply privileges across all objects within a schema in a single action rather than granting them individually for each object. This capability simplifies administration and improves database security.

Blockchain Tables

Building on the innovations from version 21c, Oracle Database 23ai further improves blockchain table functionality. These tables offer an immutable data structure, significantly strengthening the security framework of the Oracle database environment.

SQL Firewall

Oracle Database 23ai includes the Oracle SQL Firewall, a built-in security layer that provides exceptional protection directly within the database. It protects against unauthorized SQL execution and improves overall database security.

This feature provides real-time protection against common database threats, such as unauthorized SQL usage and SQL injection attacks, directly within the database. By establishing an "allow list" of approved SQL commands, the SQL Firewall quickly identifies, blocks, and logs any SQL queries that do not match the pre-approved list. Seamlessly integrated into the 23ai database architecture, the SQL Firewall ensures robust security that is difficult to bypass. Also, it is designed to have minimal to no impact on database performance.

Operations

If you manage databases, Oracle Database 23ai introduces several user-friendly enhancements designed to simplify your tasks and improve efficiency. These enhancements include:

Priority Transaction

This feature directly addresses the common challenge of transactions that block access to rows, which can disrupt applications. Oracle Database 23ai introduces a solution that automatically prioritizes transactions based on their urgency. This innovation represents a significant advancement and has been well-received by DBAs.

SQL Improvements

In addition to operational enhancements, Oracle Database 23ai expands SQL functionality with a range of features. These include the introduction of the IF NOT EXISTS syntax, a Boolean data type, and the ability to perform direct joins and updates. The update also allows for SELECT statements without a FROM clause and simplifies GROUP BY clauses with aliases. It also improves support for associative arrays and includes Unicode 14 support.

Real-time SQL Plan Management

Another significant improvement is the system's ability to detect and correct suboptimal SQL plans automatically. With real-time optimization, Oracle Database 23ai proactively maintains optimal database performance, adjusting to deliver the best possible SQL plan.

Improved Error Messages

Oracle Database 23ai has made substantial improvements to the clarity and usefulness of its error messages. This enhancement marks a significant advance towards a more efficient and user-friendly database environment. These updates reflect Oracle's commitment to advancing database technology and improving the day-to-day experience of database professionals. While improving error messages may not seem like the most dramatic upgrade, Oracle has thoroughly examined the database's most common error messages, refining and clarifying the top 90% to improve troubleshooting efficiency.

Performance

Lock-free Reservations

Locking data, explicitly or implicitly, is a fundamental aspect of database management. However, it can block business operations if data is heavily contended or if long-running transactions hold locks. Oracle Database 23ai addresses this issue with support for lock-free column value reservations. This feature enables developers to define rules when data rows need to be locked.

Read-Only Per-PDB Standby

Oracle Database 23ai introduces the ability to make pluggable database (PDB) users read-only. This configuration makes any connected session behave as if the database is in read-only mode, preventing the session from performing any write operations.

Oracle True Cache and Globally Distributed Database

Oracle Database 23ai introduces two key features to support and scale web workloads that require high scalability, availability, performance, and throughput: Oracle True Cache and Globally Distributed Database.

Oracle True Cache: Modern applications often demand massive scalability regarding the number of connections and the volume of data that can be cached. A common strategy is to place caches in front of the database, leveraging the fact that applications do not always need the most current data. For example, when a user browses a flight reservation system, the system can display flight data that is a few seconds old. However, when a flight is reserved, the system updates to show the most current data. Oracle True Cache addresses this by satisfying queries using only data from its buffer cache. Like Oracle Active Data Guard, True Cache is a fully operational read-only replica of the primary database. This allows SQL queries to be directed to the cache rather than the backend database.

Globally Distributed Database: Oracle's globally distributed database elevates your data management by distributing it across multiple databases or shards that can be hosted on different machines, whether on-premises or in the cloud. This setup effectively supports globally distributed, highly scalable, multimodel databases without requiring specialized hardware or software. It maintains all the benefits of the Oracle database, including robust consistency, comprehensive SQL capabilities, and support for structured and unstructured data, all within the familiar Oracle ecosystem. This solution also meets data sovereignty requirements and is ideal for applications that require low latency and high availability.

Oracle Database Free Get Started: https://www.oracle.com/ca-en/database/free/get-started/

Reference:

Oracle DatabaseWorld AI Edition