Manage Concurrency and Priorities on Autonomous AI Database

Concurrency and prioritization of user requests in Autonomous AI Database is determined by the database service the user is connected with, and whether compute auto scaling is enabled.

• Database Service Names for Autonomous AI Database
  Autonomous AI Database comes with predefined database services with different performance and concurrency characteristics. You must choose one of these services for each application or user that connects to the database.
• Idle Time Limits
  Autonomous AI Database has predefined idle time limits for sessions so that idle sessions do not hold system resources for a long time.
• Service Concurrency
  The consumer groups of the predefined service names provide different levels of performance and concurrency.
• Change MEDIUM Service Concurrency Limit (ECPU Compute Model)
  If your application requires customized concurrency, you can modify the concurrency limit for your Autonomous AI Database MEDIUM service.
• Change MEDIUM Service Concurrency Limit (OCPU Compute Model)
  If your application requires customized concurrency, you can modify the concurrency limit for your Autonomous AI Database MEDIUM service.
• Predefined Job Classes with Oracle Scheduler
  Autonomous AI Database includes predefined job_class values to use with Oracle Scheduler.
• Manage CPU/IO Shares on Autonomous AI Database
  Autonomous AI Database comes with predefined CPU/IO shares assigned to different consumer groups. You can modify these predefined CPU/IO shares if your workload requires different CPU/IO resource allocations.
• Manage Runaway SQL Statements on Autonomous AI Database
  Specifies how you configure Autonomous AI Database to terminate SQL statements automatically based on their query runtime or their IO usage.
• Manage Workload Resources with Database Resource Manager in Autonomous Database
  To customize resource allocation for different users, apps, or workload types in Autonomous Database, you can create and manage your own Database Resource Manager (DBRM) plans and consumer groups using cs_resource_manager subprograms.