Oracle today announced that Oracle Database 21c, the latest version of the world’s leading converged database, is available on Oracle Cloud, including the Always Free tier of Oracle Autonomous Database.
“Oracle Database 21c continues our strategy of delivering the world’s most powerful converged database engine,” said Andrew Mendelsohn, executive vice president, database server technologies, Oracle.
New Innovations in Oracle Database 21c
Oracle Database 21c is the database engine that powers Oracle database services in the cloud and on-premises, including Oracle Autonomous Database, Oracle Exadata Database Service, Oracle Exadata Database Cloud@Customer, and Oracle Exadata Database Machine.
The latest release includes more than 200 new innovations, which extend database convergence to new use cases, optimize performance, and improve developer, analyst, and data scientist productivity. Key innovations include:
Immutable Blockchain Tables: Blockchain Tables bring the key security benefits of blockchain technology to enterprise applications. Part of Oracle’s Crypto-Secure Data Management, Blockchain Tables provides immutable insert-only tables whose rows are cryptographically chained together. By providing tamper detection and prevention capabilities directly in the Oracle Database, customers can protect against illicit changes by insiders or hackers impersonating administrators or users. Blockchain Tables are part of the converged database, accessed with standard SQL, and support full analytics and transactions—making it orders of magnitude easier to use, and more functional, than existing blockchain implementations. Blockchain Tables are a free feature in all Oracle Database editions.
Native JSON Data Type: Oracle has provided powerful SQL/JSON query and indexing support for many years. Database 21c adds a new JSON data type representation, enabling up to 10x faster scans and up to 4x faster update operations. Overall, these improvements make Oracle SQL/JSON 2x faster than MongoDB and AWS DocumentDB on the YCSB benchmark. As with previous releases, users can mix or join JSON and other data types; index any JSON element for fast OLTP; use declarative parallel SQL analytics across all formats; and run complex joins across multiple JSON documents and collections—all without any need for custom application code.
AutoML for In-Database Machine Learning: Automatically builds and compares machine-learning models at scale, and facilitates the use of machine learning by non-experts. A new AutoML user interface makes it easier for non-expert users to leverage in-database machine learning. Oracle also added new algorithms for anomaly detection, regression, and deep learning analysis to our extensive library of popular, in-database machine learning algorithms.
Persistent Memory Support: Stores database data and redo logs in local Persistent Memory (PMEM), which significantly improves the performance of IO-bound workloads. SQL runs directly on data stored in the direct-mapped Persistent Memory file system, eliminating the IO code path and the need for large buffer cache. In addition, new database algorithms prevent partial or inconsistent stores to Persistent Memory.
Higher Performance Graph Models: Allows modelling of data based on relationships, and enables exploration of connections and patterns in social networks, IoT, and more. Further improvements in memory optimization reduce the amount of memory required to analyze larger graphs, which enables existing applications to run faster with no changes. In addition, users can create or extend graph algorithms using Java syntax, which can execute as native algorithms since they are compiled with the same optimizations.
Database In-Memory Automation: Oracle supports both row and column formats in the same table to allow analytics and transactions to run simultaneously on the same table. Oracle Database 21c introduces a Self-Managing In-Memory Column Store that simplifies and improves efficiency by automatically managing the placement and removal of objects in the In-Memory Column Store, then tracks usage patterns and moves and evicts objects from the column store. In addition, columns are automatically compressed based on usage patterns. Oracle Database 21c also introduces new in-memory vector join algorithms to speed up complex queries.
Sharding Automation: Native Database Sharding delivers hyperscale performance and availability while enabling global enterprises to easily meet data sovereignty and data privacy regulations. Data shards share no hardware or software, and can reside on-premises or in the cloud. To simplify the design and use of sharding, Oracle Database 21c includes a Sharding Advisor Tool that assesses a database schema plus its workload characteristics and then provides a sharded database design optimized for performance, scalability, and availability. Backup and Recovery across shards is also automated.
(Image Courtesy: www.xtech.nikkei.com)