Oracle Fleet Patching and Provisioning

About Oracle Fleet Patching and Provisioning

Oracle Fleet Patching and Provisioning (Oracle FPP) is a gold-image-based lifecycle management solution for provisioning, patching, upgrading, scaling, and maintaining Oracle Grid Infrastructure, Oracle AI Database, clusters, applications, middleware, and user-defined software. It supports Local Mode for simplified, setup-free operations and Server Mode for centrally managing large fleets across data centers. Oracle FPP provides out-of-place and rolling patching, standardization, rollback, resumable operations, scheduling, validation, dry runs, REST APIs, customizable workflows, session management, and integration with Exadata software stacks. It reduces downtime and operational risk while supporting centralized image management, deployment lineage, and automated maintenance; however, it is not supported on HP-UX or Windows, and certain standalone combined-patching scenarios are excluded.

Oracle Fleet Patching and Provisioning Features

Oracle Fleet Patching and Provisioning provides patching and upgrade capabilities for Oracle AI Database, Oracle RAC, Oracle RAC One Node, single-instance databases, Grid Infrastructure, Oracle Restart, and Exadata components, along with operating-system configuration monitoring. Its deployment models include Local Mode and a centralized Server Mode, while fleet-scale jobs, tagging, scheduling, notifications, REST APIs, configurable connectivity, and customizable authentication support large automated estates. Additional capabilities include peer image sharing, gold-image drift detection, configuration collection, external metadata repositories, arbitrary software-home provisioning, pre-checks, adaptive RAC rolling patching, combined Grid Infrastructure and database patching, batch and smart rolling updates, Application Continuity integration, custom workflows, resumable operations, and auditing.

Oracle Fleet Patching and Provisioning Architecture

Oracle Fleet Patching and Provisioning architecture consists of a highly available Oracle FPP Server, Oracle FPP Clients, and clientless targets. The Server should run on a multi-node Oracle Clusterware cluster and centrally stores gold images, working copies, client information, and metadata for users, roles, permissions, and identities. It provisions Oracle homes and other software to clients and targets through technologies such as Oracle ACFS and local file systems, supports physical and virtual servers, maintains inventories and audit records, and distributes standardized homes. Clients, integrated with Oracle Clusterware and the rhpclient component, request images and initiate operations, while clientless targets can be provisioned and managed directly by the Server. From Oracle Grid Infrastructure 26ai, the Server can use a local or external metadata repository, with separate repositories required when multiple Servers share a database container.

Oracle Fleet Patching and Provisioning Concepts

Oracle Fleet Patching and Provisioning organizes deployable software as gold images and working copies. A gold image is a reusable copy of an Oracle home that can be provisioned to new hosts and may be restricted for testing or validation before general availability. Related images can be grouped into gold image series by release, publisher, department, or other organizational criteria. A working copy is an instantiated copy of a gold image deployed on the Oracle FPP Server, an Oracle FPP Client, or a clientless target, providing the usable software home for database, Grid Infrastructure, or other software deployment.

About Oracle Update Advisor

Oracle Update Advisor automates evaluation and patch recommendations for Oracle Grid Infrastructure and Oracle AI Database homes in both Oracle FPP Local Mode and Server Mode. After user registration through Oracle Single Sign-On, it assesses a specified home or gold image and recommends patches for security, stability, and performance. Existing images return synchronous health and patch information, while custom image requests return asynchronous status details and request identifiers. Local Mode requires separate evaluation and image download steps, whereas Server Mode can evaluate, download, and import a patched image into the gold-image repository as part of the workflow.