Autonomous AI Database - Oracle AI Database Features

Describes Oracle AI Database features available with Autonomous AI Database.

Autonomous AI Database includes features that:

• Automate index management tasks, such as creating, rebuilding, and dropping indexes based on changes in the application workload.

  See Manage Automatic Indexing on Autonomous AI Database for more information.

  Note: There are restrictions for Automatic Indexing when you use JSON data with Autonomous AI Database. See SODA Notes for more information.

• Gather real-time statistics automatically while a conventional DML workload is running. Because statistics can go stale between stats gathering jobs, online statistics gathering for conventional DML helps the optimizer generate more optimal plans. Online statistics aim to reduce the possibility of the optimizer being misled by stale statistics.

  See Real-Time Statistics for more information.

• Gather statistics automatically on a more frequent basis. High-Frequency Automatic Optimizer Statistics Collection complements the standard statistics collection job. By default, the collection occurs every 15 minutes, meaning that statistics have less time in which to be stale. High-Frequency Automatic Optimizer Statistics Collection is enabled by default.

  See Configuring High-Frequency Automatic Optimizer Statistics Collection for more information.

• Quarantine execution plans for SQL statements, for example, statements that are terminated by the Resource Manager for consuming excessive system resources in an Oracle AI Database. Automatic SQL Quarantine based on Resource Manager consumption limit violations is disabled by default but any manually quarantined SQL statement will be honored.

  See Quarantine for Execution Plans for SQL Statements Consuming Excessive System Resources for more information.

• Automatically assess the opportunity for SQL plan changes to improve the performance for known statements.

  See Managing the SPM Evolve Advisor Task for more information.

• Apache ORC format is supported in Autonomous AI Database for loading and querying data in object store.

  See Create Credentials and Load Data Pump Dump Files into an Existing Table andQuery External Data with ORC, Parquet, or Avro Source Files for more information.

• Complex types are supported in Autonomous AI Database for ORC, Avro, and Parquet structured files.

  See DBMS_CLOUD Package Avro, ORC, and Parquet Complex Types for more information.