Running enterprise databases on Kubernetes calls for automation that is as dependable as the applications around them. Oracle AI Database Operator for Kubernetes extends the Kubernetes API with custom resources and controllers that help teams provision, manage, and automate the lifecycle of Oracle AI Database workloads and associated services using declarative configuration and familiar Kubernetes tooling.

Today, we’re announcing Oracle AI Database Operator for Kubernetes v2.2.0. This production release expands lifecycle automation across Oracle AI Database deployments, strengthens reconciliation and operational controls, adds new networking capabilities, and further improves security and observability.

What makes this release special

Version 2.2.0 brings meaningful operational improvements across the operator:

Together, these updates help teams standardize database operations with Kubernetes-native patterns while reducing manual steps across provisioning, maintenance, recovery, networking, and day-to-day management.

  • Expanded lifecycle automation and reconciliation for Oracle Autonomous AI Database, Oracle Autonomous AI Database on Dedicated infrastructure components, backup, restore, and multitenant workloads.
  • Broader automation for single-instance databases, Oracle Real Application Clusters, Oracle Globally Distributed AI Database, Oracle Restart, and Oracle Data Guard Broker topologies.
  • Oracle Connection Manager Controller (preview mode), supporting Oracle Connection Manager (CMAN) endpoints and generated routing rules.
  • Improved security and observability, including secure HTTPS metrics, a hardened manager security context, expanded Role Based Access Control (RBAC) and webhooks, network policies, and broader test coverage.
  • Platform compatibility across Oracle Cloud Infrastructure, Red Hat OpenShift and Oracle Kubernetes Engine (OKE), Oracle Linux Cloud Native Environment, Google Kubernetes Engine, Azure Kubernetes Service, Amazon Elastic Kubernetes Service, and Minikube.

What’s new in Oracle AI Database Operator v2.2.0

Stronger Oracle Autonomous AI Database lifecycle operations

The release deepens lifecycle automation for Oracle Autonomous AI Database and its resources. New reconciliation and validation capabilities help keep Kubernetes and OCI state aligned, while finalizer-based cleanup supports more controlled resource lifecycle management.

For backup and restore workflows, v2.2.0 adds target validation and owner references, along with point-in-time restore and OCI work-request integration. Wallet validation, wallet rotation, and backup-resource synchronization and cleanup further improve the operational foundation for these services.

More automation across database topologies

v2.2.0 expands controller capabilities across a range of database architectures:

  • Single-instance databases: Additional support for service endpoints, TCPS, Oracle True Cache, Oracle Data Guard prerequisites, clone and restore workflows, and external PVCs. Improvements also strengthen pod security, resource handling, recreation checks, and connection status.
  • Oracle Real Application Clusters (RAC) and Oracle Restart: Enhanced RAC and ASM storage lifecycle management, including disk and PVC provisioning and validation. Oracle Restart adds phased reconciliation for validation, storage, workload, and finalization, with support for static and dynamic PV/PVC handling.
  • Oracle Globally Distributed AI Database: Improved lifecycle and scaling automation for sharded topologies, including catalog and shard management and status validation.
  • Oracle True Cache: The enhancements add a new deployment mode that connects to same-cluster, cross-cluster, external databases using generated configuration blobs and ConfigMaps. It also adds service registration, TCP/TCPS connectivity, external access options, validation, lifecycle controls, samples, and documentation.
  • Data Guard Broker: Expanded topology runtime, authentication-wallet handling, validation and provisioning, Fast-Start Failover observer management, and operation tracking. Manual switchover support is now idempotent.
  • Observability: Oracle AI Database Operator for Kubernetes automates deployment, configuration, and lifecycle management of observability components alongside Oracle AI Database. Provides standardized observability across hybrid and multi-cloud deployment

These changes help teams apply repeatable, declarative workflows to database topologies that often require careful operational coordination.

Improved multitenant and ORDS operations

For multitenant deployments, the LREST and LRPDB controllers improve internal communication and configuration, including a workflow that no longer requires an HTTPS password. The release also adds application-user creation in PDBs through Kubernetes Secrets, reconciliation of PDB initialization parameters, simplified reset-bitmask status handling, and support for loading tnsnames.ora topology.

ORDS Services gain capabilities for HTTP-only edge deployments, HTTP access-log forwarding and persistence, and Instance API bootstrap. Teams can also configure resource limits, metadata, and JVM options to better align ORDS deployments with their Kubernetes operating model.

Private AI service operations

The PrivateAI controller introduces phased dependency and workload reconciliation, update-lock status, and rollout tracking. It also adds TLS-secret lifecycle support and support for vLLM and GPU-based deployments, helping teams operate private AI services with stronger Kubernetes-native controls.

Oracle Connection Manager Controller preview

Oracle AI Database Operator for Kubernetes v2.2.0 introduces the Oracle Connection Manager Controller custom resource in preview mode. It supports Oracle Connection Manager (CMAN) endpoints, generated rules, cman.ora file mode, and endpoint status. The release also adds the network.oracle.com/v4 API and the Oracle CMAN controller CRD.

This preview gives teams a new way to define and observe database networking behavior through the operator’s declarative model.

Security, observability, and operational resilience

Operator platform updates include secure HTTPS metrics, a hardened manager security context, expanded RBAC and webhook coverage, compatibility webhooks, network policy, samples, and test coverage. DatabaseObserver improvements provide safer child-resource ownership, Server-Side Apply support, and improved deployment readiness and status handling.

These enhancements are designed to make the operator more resilient and easier to operate in production environments.

Broad Kubernetes platform compatibility

Oracle AI Database Operator for Kubernetes v2.2.0 has been installed and tested on:

  • Oracle Container Engine for Kubernetes (OKE), Kubernetes
  • Red Hat OpenShift
  • Oracle Linux Cloud Native Environment
  • Google Kubernetes Engine (GKE)
  • Azure Kubernetes Service (AKS)
  • Amazon Elastic Kubernetes Service (EKS)
  • Minikube

This broad compatibility supports teams operating across OCI, on-premises environments, and multi-cloud Kubernetes estates.

Getting started

To get started, install and validate the operator, then use the controller and database-workload guides for your deployment. The project includes examples, configuration references, and operational guidance to help teams get running quickly and operate reliably in production.

For more information, visit the Oracle AI Databases for Containers and Kubernetes page.

The bottom line

Oracle AI Database Operator for Kubernetes v2.2.0 helps database and platform teams manage more of the Oracle AI Database lifecycle through the same declarative Kubernetes patterns used for their applications. With broader automation across database topologies, stronger reconciliation, improved security and observability, and a new networking preview, the release supports a more consistent and resilient operating model for Oracle AI Database on Kubernetes.