# Database Installation Guide for Linux

This file contains the text from each Database Installation Guide for Linux article landing page, release 26ai.

# Oracle AI Database Installation Checklist

Use checklists to review system requirements, and to plan and carry out Oracle AI Database installation.

Oracle recommends that you use checklists as part of your installation planning process. Using checklists can help you to confirm that your server hardware and configuration meet minimum requirements for this release, and can help you to ensure you carry out a successful installation.

## Server Hardware Checklist for Oracle AI Database Installation

Verify runlevel 3 or 5, network connectivity, and a display resolution of at least 1024 × 768 for graphical installations. Silent installations and remote graphics protocols do not require a server graphics card. Provide at least 1 GB RAM for Oracle AI Database, preferably 2 GB, and at least 8 GB for Oracle Grid Infrastructure.

Read More: [Server Hardware Checklist for Oracle AI Database Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/server-hardware-checklist-for-oracle-ai-database-installation.html)

## Operating System Checklist for Oracle AI Database Installation On Linux

Confirm that OpenSSH and a supported Linux kernel are installed, using the documented minimum versions for Linux x86-64 or supported Arm-based Oracle Linux releases. Oracle recommends using the Oracle AI Database Preinstallation RPM on Oracle Linux to configure prerequisites. Verify required packages, consult the virtualization support matrix for KVM and related technologies, and set Transparent HugePages to madvise on UEK7 and later kernels rather than disabling them.

Read More: [Operating System Checklist for Oracle AI Database Installation On Linux](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/operating-system-checklist-for-oracle-ai-database-installation-on-linux.html)

## Server Configuration Checklist for Oracle AI Database Installation

Prepare at least 1 GB in /tmp, configure swap according to available RAM and deduct HugePages memory when applicable, and establish compliant Oracle Inventory, OINSTALL groups, users, mount points, and ASCII-only Oracle home paths. Unset conflicting Oracle environment variables, configure locale settings as needed, verify that /dev/shm is a writable and executable tmpfs without noexec or nosuid restrictions, and ensure Oracle home, Oracle base, and all parent directories are not symbolic links.

Read More: [Server Configuration Checklist for Oracle AI Database Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/server-configuration-checklist-for-oracle-ai-database-installation.html)

## Oracle User Environment Configuration Checklist for Oracle AI Database Installation

Ensure the Oracle Inventory directory and OINSTALL group are available as the primary group for all Oracle software owners, creating required users, groups, resource limits, and security-specific accounts before installation. The Preinstallation RPM can create standard oracle, oinstall, and dba accounts. Unset environment variables from previous installations, set umask to 022, configure DISPLAY, and use only ASCII characters in user and group names.

Read More: [Oracle User Environment Configuration Checklist for Oracle AI Database Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/oracle-user-environment-configuration-checklist-for-oracle-ai-database-installation.html)

## Storage Checklist for Oracle AI Database Installation

Allocate at least 4.2 GB for standalone Oracle Grid Infrastructure and 8.3 GB for Oracle AI Database software on supported Linux architectures, with approximately 100 GB recommended for future patches. Select database storage on a separate file system, LVM or RAID volume, NFS/NAS file system, or Oracle ASM; install Grid Infrastructure first when using ASM. Define a recovery strategy using either a file-system fast recovery area or an ASM disk group.

Read More: [Storage Checklist for Oracle AI Database Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/storage-checklist-for-oracle-ai-database-installation.html)

## Installer Planning Checklist for Oracle AI Database

Before installation, review release notes, certification and licensing requirements, and run OUI, CVU, and current ORAchk checks to identify configuration or upgrade issues. Determine whether Grid Infrastructure is required for ASM or Oracle Restart, account for running database processes and cleanup cron jobs, and prepare My Oracle Support credentials. Decide on management tooling, memory management, root or sudo privilege delegation, and client interoperability, noting that 32-bit clients are desupported for new development in this release.

Read More: [Installer Planning Checklist for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/installer-planning-checklist-for-oracle-ai-database.html)

## Deployment Checklist for Oracle AI Database

For a single-instance deployment, install, provision through Fleet Patching and Provisioning, or clone the Oracle AI Database software; these methods can also be used when creating databases. To create a database in an existing Oracle home, use Oracle Database Configuration Assistant or Oracle Fleet Patching and Provisioning.

Read More: [Deployment Checklist for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installation-checklist/deployment-checklist-for-oracle-ai-database.html)


# Checking and Configuring the Server for Oracle AI Database

Verify that servers where you install Oracle AI Database meet the minimum requirements for installation.

This section provides minimum server requirements to complete installation of Oracle AI Database. It does not provide system resource guidelines, or other tuning guidelines for particular workloads.

## Logging In to a Remote System Using X Window System

This procedure enables Oracle Universal Installer to run on a remote server when direct GUI logins are prohibited. Start an X Window System session, authorize the remote host with `xhost`, connect to the installation server using `ssh -Y`, and switch to `root` when required; repeat the process for each installation owner account, such as `oracle` or `grid`, and consult the X server documentation if platform-specific ordering or configuration differs.

Read More: [Logging In to a Remote System Using X Window System](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/checking-and-configuring-the-server-for-oracle-ai-database/logging-in-to-a-remote-system-using-x-window-system.html)

## Setting Remote Display and X11 Forwarding Configuration

Configure the `DISPLAY` environment variable on remote terminals according to the shell in use, using values such as `hostname:0`. To prevent SSH X11 forwarding from interfering with Oracle installation, create or edit each installation owner’s `~/.ssh/config` file, set `ForwardX11 no`, and restrict `.ssh` ownership and permissions to the appropriate Oracle installation account.

Read More: [Setting Remote Display and X11 Forwarding Configuration](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/checking-and-configuring-the-server-for-oracle-ai-database/setting-remote-display-and-x11-forwarding-configuration.html)

## Checking Server Hardware and Memory Configuration

Verify server readiness by checking physical RAM, swap space, `/tmp` capacity, available memory and swap, processor architecture, and `/dev/shm` configuration with the specified system commands. Install additional memory or swap, free `/tmp` space or redirect temporary files using `TMP` and `TMPDIR`, and confirm the architecture matches the Oracle software; do not remove Oracle-specific `.oracle` directories from temporary locations after installation.

Read More: [Checking Server Hardware and Memory Configuration](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/checking-and-configuring-the-server-for-oracle-ai-database/checking-server-hardware-and-memory-configuration.html)

## Preventing Installation Errors Caused by Terminal Output Commands

During Grid Infrastructure installation, OUI uses SSH to execute commands across nodes, so hidden shell startup files must not emit terminal output. Modify each Oracle installation owner’s `.bashrc`, `.cshrc`, or equivalent files to run terminal commands such as `stty` only when standard input is a terminal; otherwise, OUI may detect the output and stop the installation.

Read More: [Preventing Installation Errors Caused by Terminal Output Commands](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/checking-and-configuring-the-server-for-oracle-ai-database/preventing-installation-errors-caused-by-terminal-output-commands.html)

## About Installing Oracle AI Database Software on Oracle Exadata Database Machine

Oracle recommends using Oracle Exadata Deployment Assistant to install Grid Infrastructure and Oracle AI Database on Exadata. For manual installations, use the Linux packages and configuration supplied with the Exadata system software without installing or upgrading RPMs or changing Linux parameters, and verify supported Oracle AI Database and Exadata software interoperability before proceeding.
Read More: [About Installing Oracle AI Database Software on Oracle Exadata Database Machine](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/checking-and-configuring-the-server-for-oracle-ai-database/about-installing-oracle-ai-database-software-on-oracle-exadata-database-machine.html)


# Automatically Configuring Oracle Linux with Oracle AI Database Preinstallation RPM

## About the Oracle AI Database Preinstallation RPM

The Oracle AI Database Preinstallation RPM simplifies operating system preparation for Oracle AI Database and Oracle Grid Infrastructure by installing required packages, resolving dependencies, creating the standard Oracle user and groups, configuring kernel and startup parameters, setting resource limits, and enabling recommended Transparent Huge Pages behavior. Install it only once during the initial software installation, do not use it on Oracle Engineered Systems, and manually configure settings for additional software owners because its limits and kernel parameters apply only to the oracle account. It installs X11 client libraries but not an X Window System server.

Read More: [About the Oracle AI Database Preinstallation RPM](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/automatically-configuring-oracle-linux-with-oracle-ai-database-preinstallation-rpm/about-the-oracle-ai-database-preinstallation-rpm.html)

## Overview of Oracle Linux Configuration with Oracle AI Database Preinstallation RPM

The Oracle AI Database Preinstallation RPM automates essential Oracle Linux preparation, including required package installation, kernel parameter configuration, and other operating system requirements. Configure it only during the first Oracle software installation and never on Oracle Engineered Systems. The overall process includes installing Oracle Linux, registering with ULN or configuring the Oracle Linux yum repository, installing and updating the preinstallation RPM, creating identically named role-based users and groups, configuring network interfaces, and preparing shared storage access for cluster nodes before installing Grid Infrastructure and the database.

Read More: [Overview of Oracle Linux Configuration with Oracle AI Database Preinstallation RPM](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/automatically-configuring-oracle-linux-with-oracle-ai-database-preinstallation-rpm/overview-of-oracle-linux-configuration-with-oracle-ai-database-preinstallation-rpm.html)

## Installing the Oracle AI Database Preinstallation RPM Using ULN

To install the preinstallation RPM through ULN, obtain the latest Oracle Linux update, register the server with ULN, and install the release-specific package as root with dnf, such as oracle-ai-database-preinstall-26ai on Oracle Linux 8, 9, or 10. The package creates standard Oracle users and groups and applies required kernel configuration; review the resulting changes in the RPM log, repeat the procedure on every cluster server, and optionally enable Ksplice when supported by the subscription.

Read More: [Installing the Oracle AI Database Preinstallation RPM Using ULN](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/automatically-configuring-oracle-linux-with-oracle-ai-database-preinstallation-rpm/installing-the-oracle-ai-database-preinstallation-rpm-using-uln.html)

## Installing Oracle AI Database Preinstallation RPM During an Oracle Linux Installation

During Oracle Linux installation, select the option to customize software, open the Servers and System administration tools package selections, and choose the Oracle AI Database Preinstallation RPM appropriate to the operating system release and architecture. Complete the installation on each cluster node so Oracle Linux automatically creates the standard Oracle installation owner and groups and configures required kernel settings. If a current package is unavailable in the installation media, install the matching RPM manually after completing the operating system installation.

Read More: [Installing Oracle AI Database Preinstallation RPM During an Oracle Linux Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/automatically-configuring-oracle-linux-with-oracle-ai-database-preinstallation-rpm/installing-oracle-ai-database-preinstallation-rpm-during-an-oracle-linux-installation.html)

## Installing Oracle Oracle AI Database Preinstallation RPM Using the Oracle Linux Yum Server

Install Oracle Linux using the standard documentation, apply current security errata and bug fixes with yum update, restart the system, and install oracle-ai-database-preinstall-26ai as root with dnf. The RPM creates the standard Oracle installation owner and groups and configures required kernel settings; environments using job-role separation must additionally create the required extended database users and groups according to their deployment needs.

Read More: [Installing Oracle Oracle AI Database Preinstallation RPM Using the Oracle Linux Yum Server](https://orahub.oci.oraclecorp.com/db-ua-dev/text-editor-poc/-/tree/no-doc-repo/automatically_configuring_oracle_linux_with_oracle_ai_database_preinstallation_rpm/26ai/installing_oracle_oracle_ai_database_preinstallation_rpm_using_the_oracle_linux_yum_server.md?ref_type=heads)

## Configure Additional Operating System Features

Before beginning installation, configure any additional operating system features required by the environment, such as IPMI or supplementary programming environments. Review the relevant preinstallation guidance and complete the configuration for each feature that will be used.
Read More: [Configure Additional Operating System Features](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/automatically-configuring-oracle-linux-with-oracle-ai-database-preinstallation-rpm/configure-additional-operating-system-features.html)


# Configuring Operating Systems for Oracle AI Database on Linux

## Guidelines for Linux Operating System Installation

Complete Linux operating system configuration and hardware checks before installing Oracle software. Ensure the deployment follows the security practices documented by the Linux distribution vendor and verify that the selected operating system is supported.

Read More: [Guidelines for Linux Operating System Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/guidelines-for-linux-operating-system-installation.html)

## Completing a Minimal or Default Linux Installation

Perform either a minimal or default Linux installation. Minimal installations require the Oracle AI Database Preinstallation RPM or manual installation of required packages, while default installations include most dependencies and reduce verification effort; Oracle recommends using default packages without customizing the RPM selection.

Read More: [Completing a Minimal or Default Linux Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/completing-a-minimal-or-default-linux-installation.html)

## About Oracle Linux with the Unbreakable Enterprise Kernel

Oracle Linux includes the Unbreakable Enterprise Kernel (UEK), which is enabled by default and optimized for Oracle database, middleware, and hardware workloads. Oracle strongly recommends UEK, although an RHEL-compatible kernel is available when strict compatibility is required. UEK is the standard kernel for Oracle products, Exadata, benchmarks, and the preinstallation RPM, while Ksplice enables live kernel updates without restarts.

Read More: [About Oracle Linux with the Unbreakable Enterprise Kernel](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/about-oracle-linux-with-the-unbreakable-enterprise-kernel.html)

## Configuring HugePages on Linux

Configure Linux HugePages to reduce page-management overhead and improve TLB performance for Oracle AI Database. Reserve at least 30 percent of system memory for standard pages and no more than 70 percent for HugePages, configure the oracle user’s memlock limit, assign its group to `vm.hugetlb_shm_group`, and persist the settings through sysctl configuration. After database creation, calculate the required HugePages value using the recommended script, set `vm.nr_hugepages`, verify allocation, and restart instances as needed.

Read More: [Configuring HugePages on Linux](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/configuring-hugepages-on-linux.html)

## Restrictions for HugePages and Transparent HugePages Configurations

Use standard HugePages for Oracle AI Database and configure Transparent HugePages as `madvise` on UEK7 and later kernels for optimal performance rather than disabling it. Transparent HugePages allocate memory dynamically through the kernel, whereas standard HugePages reserve non-swappable memory at startup; memory reserved for standard HugePages must be deducted when calculating available RAM and swap requirements.

Read More: [Restrictions for HugePages and Transparent HugePages Configurations](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/restrictions-for-hugepages-and-transparent-hugepages-configurations.html)

## Setting Transparent HugePages to madvise

On Oracle Linux and RHEL systems using UEK7 or later, set the kernel parameter `transparent_hugepage=madvise` in the GRUB configuration, regenerate the boot configuration, and restart the system. The setting can be applied immediately through the relevant sysfs entry, and its status should be verified using the appropriate Oracle Linux or RHEL Transparent HugePages path.

Read More: [Setting Transparent HugePages to madvise](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/setting-transparent-hugepages-to-madvise.html)

## About Installation Fixup Scripts

Oracle Universal Installer generates `runfixup.sh` scripts when required installation prerequisites are incomplete. These scripts can adjust kernel parameters, create and secure the central inventory directory, correct installation-owner group memberships, and set shell limits without requiring a restart. Fixup scripts do not replace complete prerequisite verification; use the installer’s integrated CVU checks or `cluvfy` for additional validation.

Read More: [About Installation Fixup Scripts](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/about-installation-fixup-scripts.html)

## Additional Drivers and Software Packages for Linux

Optional Linux components may be installed according to workload requirements, including the latest PAM library for external scheduler authentication, OCFS2 kernel and tools modules for supported Oracle Linux clusters, Messaging Gateway dependencies, current unixODBC packages for ODBC applications, and LDAP packages for directory-based features when not included in a default installation. Programming environments require supported Java, C/C++, COBOL, Unzip, and Tar versions, while web browsers are needed only for Enterprise Manager Cloud Control; OCFS2, Messaging Gateway, and some related features are not supported on Linux Arm.

Read More: [Additional Drivers and Software Packages for Linux](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/additional-drivers-and-software-packages-for-linux.html)

## Checking Kernel and Package Requirements for Linux

Verify the installed Linux distribution and release, inspect the running kernel with `uname -r` to confirm required errata, and query required RPMs with `rpm -q` or architecture-aware RPM commands. Install any missing or outdated packages and kernel updates from the Linux distributor before beginning the Oracle installation.

Read More: [Checking Kernel and Package Requirements for Linux](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/checking-kernel-and-package-requirements-for-linux.html)

## Setting Clock Source for VMs on Linux x86-64

For Linux x86-64 virtual machines, Oracle recommends using the `tsc` clock source when available. Confirm availability, set it as the active source, verify the change, and add `clocksource=tsc` to the GRUB kernel command line so the setting persists after reboot; regenerate the GRUB configuration afterward.

Read More: [Setting Clock Source for VMs on Linux x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/setting-clock-source-for-vms-on-linux-x86-64.html)

## Confirming Host Name Resolution

Ensure the installation host name resolves through DNS, NIS, or a centrally maintained `/etc/hosts` file. Use `ping` with the host name to confirm successful name resolution and network reachability before installing Oracle AI Database.

Read More: [Confirming Host Name Resolution](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/confirming-host-name-resolution.html)

## Using Automatic SSH Configuration During Installation

Cluster installations require passwordless SSH and SFTP between all member nodes so Oracle Universal Installer and configuration assistants can execute remote commands and transfer files. OUI can configure SSH automatically for the installation user, but login profiles must not produce terminal output or run commands such as `stty`; otherwise, configure SSH manually. For rare Clusterware `AttachHome` failures, set `LoginGraceTime 0` in the SSH daemon configuration on every node.

Read More: [Using Automatic SSH Configuration During Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/using-automatic-ssh-configuration-during-installation.html)

## Verifying the Disk Input Output Scheduler on Linux 8 and Later Systems

Review the disk I/O scheduler on every cluster node and consult the storage vendor for the appropriate Oracle ASM configuration. Oracle generally recommends `mq-deadline` for rotational disks and `none` for SSD or NVMe devices on supported Linux 8-or-later and SLES 15-or-later systems; verify the active scheduler through the device’s `/sys/block/.../queue/scheduler` entry.

Read More: [Verifying the Disk Input Output Scheduler on Linux 8 and Later Systems](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/verifying-the-disk-input-output-scheduler-on-linux-8-and-later-systems.html)

## Verifying Digital Signature and Integrity of Installation Archive Files

Oracle AI Database 26ai installation archives are digitally signed to verify package integrity before extraction. Use Java `jarsigner -verify` for a quick validation or add `-verbose:summary -certs` to inspect signer, certificate-chain, timestamp, digest, and signature details; a successful verification reports `jar verified.`
Read More: [Verifying Digital Signature and Integrity of Installation Archive Files](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-operating-systems-for-oracle-ai-database-on-linux/verifying-digital-signature-and-integrity-of-installation-archive-files.html)


# Operating System Requirements for x86-64 Linux Platforms

## About Operating System Requirements for x86-64 Linux Platforms

Before installing Oracle AI Database or Oracle Grid Infrastructure on x86-64 Linux, verify the supported distribution, kernel, packages, and software requirements for the intended products. Requirements may change as new platforms are certified, so consult the current certification matrix. Oracle Universal Installer checks required packages, and all cluster members must run the same operating system version unless performing an upgrade.

Read More: [About Operating System Requirements for x86-64 Linux Platforms](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/about-operating-system-requirements-for-x86-64-linux-platforms.html)

## General Guidelines for Operating System Packages

Oracle recommends using Oracle Linux with the Oracle AI Database Preinstallation RPM, which configures the operating system and installs required dependencies. Interactive installation requires an X Window System package such as libX, while compiler packages such as gcc and gcc-c++ are not required. These requirements do not apply to Oracle Engineered Systems, which provide integrated operating system software and required packages.

Read More: [General Guidelines for Operating System Packages](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/general-guidelines-for-operating-system-packages.html)

## Supported Oracle Linux 10 Distributions for x86-64

Oracle Linux 10 requires OpenSSH and, beginning with Oracle AI Database 26ai Release Update 23.26.2, supports UEK 8 kernel 6.12.0-104.43.4.2.el10uek.x86\_64 or later and Red Hat Compatible Kernel 6.12.0-55.40.1.0.1.el10\_0.x86\_64 or later. Install current required packages through the Oracle AI Database Preinstallation RPM or appropriate repositories; optional packages include ipmiutil, net-tools, nfs-utils, and client-related libnsl packages. Review applicable patches and known issues before installation.

Read More: [Supported Oracle Linux 10 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-oracle-linux-10-distributions-for-x86-64.html)

## Supported Oracle Linux 9 Distributions for x86-64

Oracle Linux 9 requires OpenSSH and supports Oracle Linux 9.2 with UEK 7 kernel 5.15.0-201.135.6.el9uek.x86\_64 or later, Red Hat Compatible Kernel 5.14.0-284.30.1.el9\_2.x86\_64 or later, and, beginning with Release Update 23.26.2, Oracle Linux 9.6 with UEK 8 kernel 6.12.0-1.23.3.2.el9uek.x86\_64 or later. Use the Oracle AI Database Preinstallation RPM to install required packages and configuration, with optional packages available for IPMI, RAC, ACFS, and Database Client deployments.

Read More: [Supported Oracle Linux 9 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-oracle-linux-9-distributions-for-x86-64.html)

## Supported Oracle Linux 8 Distributions for x86-64

Oracle Linux 8 requires OpenSSH and supports version 8.6 with UEK 6 kernel 5.4.17-2136.312.3.4.el8uek.x86\_64 or later, Red Hat Compatible Kernel 4.18.0-372.26.1.0.1.el8\_6.x86\_64 or later, and, beginning with Release Update 23.5, version 8.8 with UEK 7 kernel 5.15.0-202.135.2.el8uek.x86\_64 or later. Install current dependencies using the Oracle AI Database Preinstallation RPM or vendor repositories; optional packages support IPMI, Database Client, RAC, and ACFS requirements.

Read More: [Supported Oracle Linux 8 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-oracle-linux-8-distributions-for-x86-64.html)

## Supported Red Hat Enterprise Linux 10 Distributions for x86-64

Red Hat Enterprise Linux 10 requires OpenSSH and, beginning with Oracle AI Database 26ai Release Update 23.26.2, kernel 6.12.0-55.40.1.0.1.el10\_0.x86\_64 or later. Install the specified current packages, including libraries, development components, X Window System libraries, security utilities, monitoring tools, and make, from enabled Red Hat repositories. Optional packages include ipmiutil, net-tools, and nfs-utils; review relevant installation patches and known issues.

Read More: [Supported Red Hat Enterprise Linux 10 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-red-hat-enterprise-linux-10-distributions-for-x86-64.html)

## Supported Red Hat Enterprise Linux 9 Distributions for x86-64

Red Hat Enterprise Linux 9 requires OpenSSH and version 9.2 with kernel 5.14.0-284.30.1.el9\_2.x86\_64 or later. Required packages include core system libraries, compatibility OpenSSL, X libraries, compiler and RDMA components, virtualization libraries, policy utilities, monitoring tools, and development packages. Optional packages support IPMI, Database Client, RAC, and ACFS installations.

Read More: [Supported Red Hat Enterprise Linux 9 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-red-hat-enterprise-linux-9-distributions-for-x86-64.html)

## Supported Red Hat Enterprise Linux 8 Distributions for x86-64

Red Hat Enterprise Linux 8 requires OpenSSH and version 8.6 with kernel 4.18.0-372.26.1.0.1.el8\_6.x86\_64 or later. Install the required system, compiler, X Window System, RDMA, security, monitoring, and development packages from vendor repositories. Optional packages include ipmiutil, Database Client libnsl packages, net-tools for RAC and Clusterware, and nfs-utils for ACFS.

Read More: [Supported Red Hat Enterprise Linux 8 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-red-hat-enterprise-linux-8-distributions-for-x86-64.html)

## Supported SUSE Linux Enterprise Server 15 Distributions for x86-64

SUSE Linux Enterprise Server 15 requires OpenSSH and SP5 with kernel 5.14.21-150500.53-default or later. Install the listed current packages, including development libraries, X libraries, capability and ELF libraries, compilers, make, policy utilities, monitoring tools, and related runtime dependencies. Optional packages are ipmiutil for IPMI and net-tools for Oracle RAC and Clusterware.

Read More: [Supported SUSE Linux Enterprise Server 15 Distributions for x86-64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/supported-suse-linux-enterprise-server-15-distributions-for-x86-64.html)

## Installing Operating System Packages

Install the latest operating system packages from the appropriate vendor repository using DNF or YUM on Oracle Linux and Red Hat Enterprise Linux, or YaST on SUSE Linux Enterprise Server, after enabling the required channel or repository. For example, install bc with dnf install bc or yum install bc on Oracle Linux and Red Hat Enterprise Linux, or with yast --install bc on SUSE Linux Enterprise Server.
Read More: [Installing Operating System Packages](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-x86-64-linux-platforms/installing-operating-system-packages.html)


# Operating System Requirements for Linux for Arm (aarch64)

## About Operating System Requirements for Linux for Arm (aarch64)

Before installing Oracle AI Database or Oracle Grid Infrastructure on Linux for Arm (aarch64), verify that the supported operating system distribution, kernel, packages, and software prerequisites are installed. Oracle Universal Installer checks these requirements, which should be validated in advance; current hardware and operating system certification should also be confirmed through My Oracle Support. All cluster members must use the same operating system version, except during an operating system upgrade.

Read More: [About Operating System Requirements for Linux for Arm aarch64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-linux-for-arm-aarch64/about-operating-system-requirements-for-linux-for-arm-aarch64.html)

## Supported Oracle Linux 9 Distributions on Linux for Arm (aarch64)

Oracle AI Database 26ai supports Oracle Linux 9 on Linux for Arm beginning with Release Update 23.26.2, requiring Oracle Linux 9.4 or later with Unbreakable Enterprise Kernel 7 version 5.15.0-307.178.5.el9uek.aarch64 or later. Oracle recommends using the latest Oracle Linux 9 release. The oracle-ai-database-preinstall-26ai RPM, obtained through the Oracle Linux 9 Unbreakable Linux Network channel or Oracle Linux yum repository, installs the required database and Grid Infrastructure packages, dependencies, and system configuration, including compiler, runtime, storage, X11, security, monitoring, and related libraries.

Read More: [Supported Oracle Linux 9 Distributions on Linux for Arm aarch64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-linux-for-arm-aarch64/supported-oracle-linux-9-distributions-on-linux-for-arm-aarch64.html)

## Supported Oracle Linux 8 Distributions on Linux for Arm (aarch64)

Oracle AI Database 26ai supports Oracle Linux 8 on Linux for Arm beginning with Release Update 23.7, requiring Oracle Linux 8.6 or later with Unbreakable Enterprise Kernel 7 version 5.15.0-205.149.5.1.el8uek.aarch64 or later. Oracle recommends using the latest Oracle Linux 8 release. The oracle-ai-database-preinstall-26ai RPM, available through the Oracle Linux 8 Unbreakable Linux Network channel or Oracle Linux yum repository, automatically installs the required Grid Infrastructure and database packages, dependencies, and system settings, including development tools, runtime libraries, storage and networking libraries, X11 components, security utilities, monitoring tools, and related dependencies.
Read More: [Supported Oracle Linux 8 Distributions on Linux for Arm aarch64](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/operating-system-requirements-for-linux-for-arm-aarch64/supported-oracle-linux-8-distributions-on-linux-for-arm-aarch64.html)


# Configuring Users Groups and Environments for Oracle Grid Infrastructure and Oracle AI Database

Before installation, create operating system groups and users, and configure user environments.

## Required Operating System Groups and Users

Oracle installations require an installation owner, a shared Oracle Inventory group as the primary group for all installation owners, and one or more operating system privilege groups. Verify whether an existing central inventory and `oraInst.loc` file exist, reuse the same inventory for all installations, and ensure installation owners can write to it. Create the inventory group if needed, and use designated `grid` and `oracle` owners when separating Grid Infrastructure and database administration. Installation owners must use the inventory group as their primary group and appropriate database, ASM, backup, Data Guard, encryption, and RAC groups as secondary groups; for Oracle Restart, ensure `grid` belongs to `racdba`.

Read More: [Required Operating System Groups and Users](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-users-groups-and-environments-for-oracle-grid-infrastructure-and-oracle-ai-database/required-operating-system-groups-and-users.html)

## Oracle Installations with Standard and Job Role Separation Groups and Users

Oracle supports standard and job role separation configurations through operating system users and groups that control administrative privileges for Grid Infrastructure, ASM, and databases. A single owner and group may provide all privileges, but Oracle recommends separate inventory and administrative groups when role separation is required. Standard database groups include OSDBA (`dba`) and optional OSOPER (`oper`), while extended groups provide limited SYSBACKUP, SYSDG, SYSKM, and SYSRAC privileges. ASM administration uses OSASM (`asmadmin`), optional OSOPER for ASM (`asmoper`), and OSDBA for ASM (`asmdba`); the `ASMSNMP` account can be created with limited monitoring privileges for Enterprise Manager.

Read More: [Oracle Installations with Standard and Job Role Separation Groups and Users](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-users-groups-and-environments-for-oracle-grid-infrastructure-and-oracle-ai-database/oracle-installations-with-standard-and-job-role-separation-groups-and-users.html)

## Creating Operating System Privileges Groups

Create the operating system groups required by the installation when they do not already exist, using conventional names such as `asmadmin` for OSASM, `asmdba` for ASM disk access, `asmoper` for ASM startup and shutdown, `dba` for database SYSDBA, and `oper` for database SYSOPER. Database installations also require groups for SYSBACKUP (`backupdba`), SYSDG (`dgdba`), SYSKM (`kmdba`), and SYSRAC (`racdba`). Group identifiers should be selected according to the operating system’s existing allocation and organizational policies.

Read More: [Creating Operating System Privileges Groups](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-users-groups-and-environments-for-oracle-grid-infrastructure-and-oracle-ai-database/creating-operating-system-privileges-groups.html)

## Creating Operating System Oracle Installation User Accounts

Create Oracle software owner accounts before installation, typically `oracle` for database software and `grid` for Grid Infrastructure, assigning the Oracle Inventory group as the primary group and the required privilege groups as secondary groups. Record each user ID for preinstallation tasks. If an existing owner lacks the required memberships, update its groups before installation, but do not change the primary inventory group or use different primary groups among owners because this can corrupt the central inventory.

Read More: [Creating Operating System Oracle Installation User Accounts](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-users-groups-and-environments-for-oracle-grid-infrastructure-and-oracle-ai-database/creating-operating-system-oracle-installation-user-accounts.html)

## Configuring Software Owner User Environments

Before installation, configure each `oracle` or `grid` environment with `umask 022`, suitable file-descriptor and process limits, and a valid `DISPLAY` setting for graphical installers. Remove conflicting Oracle variables such as `ORACLE_HOME`, `ORACLE_SID`, `ORACLE_BASE`, `ORA_NLS10`, `TNS_ADMIN`, and `ORA_CRS_HOME`, and remove Oracle home paths from `PATH`. Ensure temporary storage provides at least 1 GB, using a nonshared directory for RAC installations, and verify the resulting environment. Recommended limits include at least 1,024 soft and 65,536 hard open files, 2,047 soft and 16,384 hard processes, a stack size of approximately 10,240–32,768 KB, and sufficient locked memory for HugePages or database operation; re-login after changing limits.

Read More: [Configuring Software Owner User Environments](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-users-groups-and-environments-for-oracle-grid-infrastructure-and-oracle-ai-database/configuring-software-owner-user-environments.html)

## Creating Oracle Database Vault User Accounts

If Oracle Database Vault is enabled by default, create and configure its required accounts before use. The Database Vault Owner account is mandatory, while a separate Account Manager account may also be created for administration. Database Vault establishes baseline auditing for security configuration, catalog information, system privileges, and Oracle Label Security, and initializes related database parameters with default values.
Read More: [Creating Oracle Database Vault User Accounts](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-users-groups-and-environments-for-oracle-grid-infrastructure-and-oracle-ai-database/creating-oracle-database-vault-user-accounts.html)


# Configuring Networks for Oracle AI Database

## About Oracle AI Database Network Configuration Options

Oracle AI Database requires no special network configuration on a server with a static, resolvable host name, IP address, and at least one network interface. Special considerations apply to multihomed, multialias, or temporarily non-networked computers, including ensuring firewall ports are open and configuring connectivity or database services as needed. Network interfaces and switches should support at least 1 GbE, with 25 GbE or higher recommended for NVMe-oF storage.

Read More: [About Oracle AI Database Network Configuration Options](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-networks-for-oracle-ai-database/about-oracle-ai-database-network-configuration-options.html)

## About Assigning Global Database Names During Installation

During installation, the database name defines the DB\_NAME, DB\_UNIQUE\_NAME, and DB\_DOMAIN values. The global database name combines the database name and domain to uniquely identify the database on a network, while the SID uniquely identifies a database instance on a host. Database and domain names have defined length and character restrictions, and DB\_NAME does not have to be the first eight characters of DB\_UNIQUE\_NAME.

Read More: [About Assigning Global Database Names During Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-networks-for-oracle-ai-database/about-assigning-global-database-names-during-installation.html)

## Network Configuration for Computers Completed After Installation

Oracle AI Database can be installed on computers without fixed network addresses, including DHCP-configured systems, but the computer must be able to resolve and reach itself. Before installation, use ping with both the host name and fully qualified host name, verifying that the expected local IP address is returned and that the fully qualified name is defined in /etc/hosts. After networking is enabled, the system can use either static IP addressing or DHCP.

Read More: [Network Configuration for Computers Completed After Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-networks-for-oracle-ai-database/network-configuration-for-computers-completed-after-installation.html)

## Network Configuration for Multihome Computers

Installing Oracle AI Database on a multihomed computer requires setting ORACLE\_HOSTNAME to identify the appropriate host name or interface. Oracle Universal Installer uses this value when available, or otherwise derives host information from /etc/hosts and installation inputs. Clients must resolve and ping both the host name and fully qualified domain name associated with the database instance or alias.

Read More: [Network Configuration for Multihome Computers](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-networks-for-oracle-ai-database/network-configuration-for-multihome-computers.html)

## Setting the ORACLE\_HOSTNAME Environment Variable

Set ORACLE\_HOSTNAME as the Oracle installation user before installation, using export in Bourne, Bash, or Korn shells or setenv in the C shell. The value should be the fully qualified host name intended for the Oracle AI Database installation.

Read More: [Setting the ORACLE\_HOSTNAME Environment Variable](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-networks-for-oracle-ai-database/setting-the-oracle_hostname-environment-variable.html)

## Network Configuration for Computers with Multiple Aliases

Installing Oracle AI Database on a computer with multiple host aliases requires setting ORACLE\_HOSTNAME to the preferred host name. These aliases resolve to the same IP address through the naming service, so the environment variable determines which computer name Oracle uses during installation.
Read More: [Network Configuration for Computers with Multiple Aliases](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-networks-for-oracle-ai-database/network-configuration-for-computers-with-multiple-aliases.html)


# Supported Storage Options for Oracle AI Database and Oracle Grid Infrastructure

Review supported storage options as part of your installation planning process.

## Supported Storage Options for Oracle AI Database

Oracle AI Database supports Oracle ASM, Oracle ACFS, local file systems, NFS on NAS, and NVMe-oF for different file types. ASM and NVMe-oF support data and recovery files but not binaries, while ACFS, local file systems, and NFS support binaries; local file systems are not recommended for data or recovery files. Use dedicated ASM disks, avoid loopback, raw, and block devices, and select combinations that meet all storage requirements. ASM can also store Oracle Clusterware files.

Read More: [Supported Storage Options for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/supported-storage-options-for-oracle-ai-database.html)

## About Oracle Grid Infrastructure for a Standalone Server

Oracle Grid Infrastructure for a standalone server combines Oracle Restart and Oracle ASM for single-instance databases, providing volume management, file-system services, monitoring, and automatic restart. Oracle Restart starts components in the correct order, monitors their health, and recovers failed database instances, listeners, services, and ASM instances. It must be installed before using ASM unless installed and registered afterward, is limited to nonclustered environments, and cannot be used on cluster-member nodes; clustered systems require Grid Infrastructure for a Cluster.

Read More: [About Oracle Grid Infrastructure for a Standalone Server](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/about-oracle-grid-infrastructure-for-a-standalone-server.html)

## Oracle ACFS and Oracle ADVM

Oracle ACFS extends ASM to provide general-purpose file-system support for single-instance and clustered environments, while Oracle ADVM supplies volume management and the device interface through which ACFS accesses ASM. ACFS and ADVM support specified Oracle Linux, Red Hat Enterprise Linux, and SUSE Linux Enterprise Server kernels, but not Linux Containers; kernel patches and appropriate SELinux configuration may be required. ACFS requires ASM, can store binaries and administrative files, and has restrictions on database-file encryption, replication, Clusterware files, Oracle Restart resources, and data files, particularly in standalone Oracle Restart configurations.

Read More: [Oracle ACFS and Oracle ADVM](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/oracle-acfs-and-oracle-advm.html)

## File System Options for Oracle AI Database

Oracle AI Database files can be created on physically attached disks, LVM or RAID volumes, or NFS mounted from NAS, with Direct NFS available to simplify administration and improve performance. Oracle recommends separating database storage from operating-system and Oracle software file systems, distributing files across disks when appropriate, and using SAME for striped and mirrored storage. Oracle Managed Files can further automate file naming and management by allowing database object names instead of explicit file names.

Read More: [File System Options for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/file-system-options-for-oracle-ai-database.html)

## Guidelines for Placing Oracle AI Database Files On a File System or Logical Volume

Database file systems should use storage dedicated to the database and may consist of one or multiple file systems. A single file system should preferably reside on RAID or a multi-device logical volume using SAME, while multiple file systems should use separate physical devices to distribute I/O, separate control files, and support OFA practices. Installation requires at least 2 GB of free space for a preconfigured database, with greater capacity planned for production, and the installation user must have write access to the selected paths.

Read More: [Guidelines for Placing Oracle AI Database Files On a File System or Logical Volume](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/guidelines-for-placing-oracle-ai-database-files-on-a-file-system-or-logical-volume.html)

## About NFS Storage for Data Files

NFS on NAS provides shared file access suitable for both single-instance and Oracle RAC databases. Database performance depends heavily on the network and NAS infrastructure, so dedicated private connections using Gigabit Ethernet or faster are recommended. Before installation, all required NFS file systems must be correctly configured, mounted, and available to the relevant servers.

Read More: [About NFS Storage for Data Files](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/about-nfs-storage-for-data-files.html)

## About Direct NFS Client Mounts to NFS Storage Devices

Direct NFS Client integrates NFS access into Oracle software to optimize database I/O, supports NFSv3, NFSv4, NFSv4.1, and pNFS, and can use a dispatcher to improve scalability for large or parallel NFS deployments. NFS servers require a write size of at least 32768, and mount points must be available to both the operating-system NFS client and dNFS for fallback support. dNFS searches configured mount files in a defined order, permits only one active NFS client implementation per instance, defaults to NFSv3 when unspecified, and maps unsupported NFSv4.2 requests to NFSv4.1.
Read More: [About Direct NFS Client Mounts to NFS Storage Devices](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/supported-storage-options-for-oracle-ai-database-and-oracle-grid-infrastructure/about-direct-nfs-client-mounts-to-nfs-storage-devices.html)


# Configuring File System Storage for Oracle AI Database

## About File System Storage for Oracle AI Database

This page outlines procedures for using file system storage with Oracle AI Database and recommends placing database data files and Oracle software on separate disks for organization and performance. It also notes that Oracle Persistent Memory Database and Oracle Memory Speed File System are deprecated in release 26ai because Intel is discontinuing Optane Persistent Memory hardware.

Read More: [About File System Storage for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-file-system-storage-for-oracle-ai-database/about-file-system-storage-for-oracle-ai-database.html)

## Deploying Oracle Direct NFS for Oracle AI Database

This page describes deploying Oracle Direct NFS for databases using NFS storage, including recommended separation of database files and software, NFS buffer settings of at least 32768, TCP buffer tuning for network performance, and creation of an `oranfstab` configuration file. It explains server, network path, export, mount, timeout, protocol version, routing, management, community, and security attributes, including Kerberos options, and provides configuration examples for multiple paths, exports, and authentication modes.

Read More: [Deploying Oracle Direct NFS for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-file-system-storage-for-oracle-ai-database/deploying-oracle-direct-nfs-for-oracle-ai-database.html)

## Enabling and Disabling Direct NFS Client Control of NFS

This page explains how to enable or disable Direct NFS Client control of NFS using the `dnfs_on` and `dnfs_off` make targets after configuring `oranfstab`. Disabling requires removing the configuration file, and databases using removed NFS paths must be restarted. It also describes enabling Hybrid Columnar Compression by verifying SNMP on the ZFS storage server, configuring management and community settings when needed, and ensuring the Net-SNMP library is installed.

Read More: [Enabling and Disabling Direct NFS Client Control of NFS](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-file-system-storage-for-oracle-ai-database/enabling-and-disabling-direct-nfs-client-control-of-nfs.html)

## Deploying a DAX-Enabled File System for Persistent Memory Database

This page explains how to deploy an XFS file system mounted with DAX for Persistent Memory Database, while noting that OMS File System is deprecated in Oracle AI Database 26ai because Intel is discontinuing Optane PMEM. The process requires an Intel PMEM device in `fsdax` mode, a DAX-capable XFS file system, an instance started in `NOMOUNT` mode, creation of a PMEM filestore, and database or tablespace creation. Recommended configuration includes 2 MB alignment, separate metadata and data partitions, 2 MB XFS extents, and mounting with appropriate DAX and real-time device options.

Read More: [Deploying a DAX-Enabled File System for Persistent Memory Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-file-system-storage-for-oracle-ai-database/deploying-a-dax-enabled-file-system-for-persistent-memory-database.html)

## Creating a Database on PMEM Storage Using Oracle DBCA

This page describes using Oracle DBCA in interactive or silent mode to create a single-instance Oracle AI Database with PMEM storage. The workflow covers launching DBCA, selecting advanced configuration and a single-instance deployment, defining the global database name, SID, CDB and PDB options, and selecting PMEM File System as the database storage type. Users must specify the PMEM location and size, may enable Oracle Managed Files, and then complete the remaining configuration prompts.
Read More: [Creating a Database on PMEM Storage Using Oracle DBCA](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-file-system-storage-for-oracle-ai-database/creating-a-database-on-pmem-storage-using-oracle-dbca.html)


# Configuring Storage for Oracle Grid Infrastructure for a Standalone Server

## About Configuring Storage for Oracle Grid Infrastructure for a Standalone Server

Oracle Grid Infrastructure for a standalone server, also known as Oracle Restart, supports single-instance Oracle AI Database deployments and includes Oracle ASM for volume management, database files, recovery files, binaries, and general-purpose file systems. Oracle Grid Infrastructure must be installed before creating the database when using Oracle ASM or Oracle Restart; Oracle Restart cannot be installed on or added to a cluster member node.

Read More: [About Configuring Storage for Oracle Grid Infrastructure for a Standalone Server](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server/about-configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server.html)

## Configuring Storage for Oracle Automatic Storage Management

Configuring Oracle ASM requires determining storage usage, disk capacity, redundancy, allocation unit size, and failure-group design. External redundancy relies on storage-level protection, normal redundancy generally provides two-way mirroring, high redundancy provides three-way mirroring, flex redundancy supports file-specific protection and quotas, and extended redundancy protects across sites. Disk groups should use appropriately sized, similarly performing devices owned by the installation user, with separate groups recommended for database and recovery files. Existing groups can be inspected with ASM tools, NAS files can serve as ASM devices when correctly created and permissioned, and DAS or SAN disks should generally have one partition per physical disk.

Read More: [Configuring Storage for Oracle Automatic Storage Management](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server/configuring-storage-for-oracle-automatic-storage-management.html)

## Configuring Storage Device Path Persistence Using Oracle ASMLIB

Oracle ASMLIB simplifies ASM device management by providing persistent disk labels and permissions across system restarts and operating-system upgrades. Configure the disk discovery string as `ORCL:*`, and use an interactive installation method that runs ASMCA when creating a database with the ASM library driver. Install the appropriate ASMLIB packages for the operating system and kernel, using ASMLIB v3 on Oracle Linux 8 and later kernels; ASMLIB is not supported on IBM Linux on System z.

Read More: [Configuring Storage Device Path Persistence Using Oracle ASMLIB](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server/configuring-storage-device-path-persistence-using-oracle-asmlib.html)

## Deinstalling Oracle ASMLIB On Oracle AI Database

To remove unused Oracle ASMLIB, log in as root, stop the database, ASM, and Oracle Restart stack, unload and disable ASMLIB, remove the `oracleasm-support` and `oracleasmlib` RPMs, and verify and remove any remaining related packages or configuration files. After removal, restart the Oracle Restart stack.

Read More: [Deinstalling Oracle ASMLIB On Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server/deinstalling-oracle-asmlib-on-oracle-ai-database.html)

## Creating Directories for Oracle AI Database Files

Database and recovery files can reside on file systems separate from the Oracle base directory. Use `df -h` to identify suitable mount points, allowing at least 1.5 GB for database files on one file system or 3.5 GB across multiple file systems and at least 2 GB for recovery files, adding requirements when sharing storage. If the installer lacks write access, manually create the directories, assign the Oracle installation owner and inventory group, set appropriate permissions, and preferably place the fast recovery area on a separate physical disk from database files.

Read More: [Creating Directories for Oracle AI Database Files](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server/creating-directories-for-oracle-ai-database-files.html)

## Configuring NVMe Over Fabrics Storage Devices with Oracle ASM

Oracle AI Database 26ai supports Oracle-native user-space NVMe over Fabrics over TCP for ASM disk groups on Linux, using Linux kernel 5.4 or later, `nvmet_tcp` targets, and dedicated 25 GbE-or-faster networking. Configure target subsystems, namespaces, ports, and unique NQNs with `nvmetcli`; build the NVMe ODM library; create `/etc/oracle/nvm/nvmip.ora` on each Grid Infrastructure node with initiator and target addresses; and discover devices through `asm_diskstring='n:/*/*'`. All standard ASM storage and failure-domain guidelines apply, and every initiator must communicate with all configured targets.
Read More: [Configuring NVMe Over Fabrics Storage Devices with Oracle ASM](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-storage-for-oracle-grid-infrastructure-for-a-standalone-server/configuring-nvme-over-fabrics-storage-devices-with-oracle-asm.html)


# Installing and Configuring Oracle Grid Infrastructure for a Standalone Server

## About Installing Oracle Grid Infrastructure for a Standalone Server

Oracle Grid Infrastructure for a standalone server combines Oracle Restart and Oracle ASM. Databases created after installation are automatically registered with Oracle Restart, while existing databases, listeners, ASM, and related components must be registered manually. Oracle Restart supports multiple single-instance databases, and starting with Oracle AI Database 26ai, `gridSetup.sh` replaces the desupported `config.sh` utility.

Read More: [About Installing Oracle Grid Infrastructure for a Standalone Server](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/about-installing-oracle-grid-infrastructure-for-a-standalone-server.html)

## About Image-Based Oracle Grid Infrastructure Installation

Image-based installation simplifies Oracle Grid Infrastructure deployment by extracting the software image into a newly created Grid home and running `gridSetup.sh`. It supports cluster configuration, standalone-server Oracle Restart configuration, upgrades, software-only installations, node changes, silent deployments, and customized patched images, while supporting GUI, silent, and Fleet Patching and Provisioning workflows. Grid home paths must follow Oracle Optimal Flexible Architecture recommendations.

Read More: [About Image-Based Oracle Grid Infrastructure Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/about-image-based-oracle-grid-infrastructure-installation.html)

## Setup Wizard Installation Options for Creating Images

The `runInstaller` and `gridSetup.sh` setup wizards support gold-image creation through options that specify the destination, excluded files, and help output. The `-createGoldImage` option creates an image from the current Oracle home, `-destinationLocation` defines the output directory, and `-exclFiles` excludes specified files or paths, enabling reusable database or Grid Infrastructure deployment images.

Read More: [Setup Wizard Installation Options for Creating Images](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/setup-wizard-installation-options-for-creating-images.html)

## Installing Oracle Grid Infrastructure for a Standalone Server

Install Oracle Restart and Oracle ASM by preparing the required users, groups, network, storage, Oracle base, and inventory directories; extracting the current Grid Infrastructure gold image into an OFA-compliant, non-symbolic-link Grid home; and running `gridSetup.sh`. Select the standalone-server configuration, configure the default or additional ASM disk groups, provide or automate root-script execution, and then install the database using ASM storage. ASMFD is desupported in Oracle AI Database 26ai, with ASMlib v3 recommended for applicable Linux environments.

Read More: [Installing Oracle Grid Infrastructure for a Standalone Server](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/installing-oracle-grid-infrastructure-for-a-standalone-server.html)

## Installing Oracle Grid Infrastructure for a Standalone Server for an Existing Database

To manage an existing database with Oracle Restart, stop its listeners, install Grid Infrastructure in a separate Grid home, restart the listeners, and register the database and listeners with `srvctl`. Oracle Restart can manage databases from the same release and up to one release lower, while earlier releases may coexist without being managed by Oracle Restart.

Read More: [Installing Oracle Grid Infrastructure for a Standalone Server for an Existing Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/installing-oracle-grid-infrastructure-for-a-standalone-server-for-an-existing-database.html)

## Installing Oracle Grid Infrastructure for a Standalone Server Using a Software-Only Installation

A software-only installation places Grid Infrastructure binaries in the Grid home without validating or configuring Oracle Restart, so it is intended for advanced users. The process involves extracting the image, verifying prerequisites with `runcluvfy.sh`, running `gridSetup.sh` with Set Up Software Only, executing prompted root scripts, and then completing manual configuration with ASM or, without ASM, `roothas.sh` followed by listener registration using `srvctl` and Oracle Network Configuration Assistant. This method does not support upgrades from previous Oracle Restart releases.

Read More: [Installing Oracle Grid Infrastructure for a Standalone Server Using a Software-Only Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/installing-oracle-grid-infrastructure-for-a-standalone-server-using-a-software-only-installation.html)

## Installing Oracle Grid Infrastructure for a Standalone Server Using Command-Line Interface

Starting with Oracle Grid Infrastructure 26ai, standalone-server installation and configuration can be performed from the command line. After preparing OFA-compliant directories and extracting the image into a suitable Grid home, run `gridSetup.sh -silent -configureStandaloneServer` with the required operating-system groups, Oracle base, inventory, disk group, disk, redundancy, allocation-unit, discovery, and root-script options, then provide the root password when prompted. ASMFD remains desupported in this release.

Read More: [Installing Oracle Grid Infrastructure for a Standalone Server Using Command-Line Interface](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/installing-oracle-grid-infrastructure-for-a-standalone-server-using-command-line-interface.html)

## Testing the Oracle Automatic Storage Management Installation

Test the ASM installation by setting `ORACLE_SID` to the ASM instance, setting `ORACLE_HOME` to the Grid home, and running `asmcmd lsdg` to display disk groups. If ASM is not running, start it with `srvctl start asm`, after which ASMCMD can connect using SYSASM privileges.

Read More: [Testing the Oracle Automatic Storage Management Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/testing-the-oracle-automatic-storage-management-installation.html)

## Relinking Oracle Restart and Oracle ASM Binaries

Relink Oracle Restart and ASM binaries after operating-system patching or upgrades, first stopping all executables and applications using the Grid home or its shared libraries. As root, unlock the Grid home with `roothas.sh -unlock`; as the `grid` user, run the Grid home `relink` utility; then as root, run `rootadd_rdbms.sh` and relock the home with `roothas.sh -lock`.

Read More: [Relinking Oracle Restart and Oracle ASM Binaries](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/relinking-oracle-restart-and-oracle-asm-binaries.html)

## Configuring Oracle ASM Disk Groups Manually using Oracle ASMCA

After standalone-server installation, ASMCA can create and configure additional ASM disk groups, Oracle ADVM, and Oracle ACFS. If no ASM instance exists, the installation process creates one; afterward, launch ASMCA from the Grid home’s `bin` directory to manually configure ASM storage and related services.

Read More: [Configuring Oracle ASM Disk Groups Manually using Oracle ASMCA](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/configuring-oracle-asm-disk-groups-manually-using-oracle-asmca.html)

## Enabling Oracle ACFS on Oracle Restart Configurations

Enable root access for Oracle ACFS by logging in as root, changing to the Grid home’s `crs/install` directory, and running `roothas.sh -lockacfs`.

Read More: [Enabling Oracle ACFS on Oracle Restart Configurations](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/enabling-oracle-acfs-on-oracle-restart-configurations.html)

## Applying Patches During an Oracle Grid Infrastructure Installation or Upgrade

Obtain the required Release Updates or one-off patches, place them in an accessible directory, change to the Grid home, and run `gridSetup.sh` with `-applyRU` and/or `-applyOneOffs`. Patches may be applied separately or together, after which the remaining setup or upgrade wizard steps must be completed.

Read More: [Applying Patches During an Oracle Grid Infrastructure Installation or Upgrade](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/applying-patches-during-an-oracle-grid-infrastructure-installation-or-upgrade.html)

## Patching and Switching Oracle Grid Infrastructure Homes

Use `switchGridHome` to move Oracle Grid Infrastructure between homes containing different release updates or one-off patches. Create a distinct new Grid home, initialize it with the source home’s configuration, apply the required patches, and switch to it in silent or interactive mode; services then run from the new home. To revert, unlock the old home as root and run its `switchGridHome` command as the `grid` user, optionally using zero-downtime or driver-update options.

Read More: [Patching and Switching Oracle Grid Infrastructure Homes](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/patching-and-switching-oracle-grid-infrastructure-homes.html)

## Changing the Oracle Grid Infrastructure Base Path

To change the Grid Infrastructure base directory, create and own the new base, stop Oracle Restart and TFA, copy the Grid files, preserve the old base as a backup, and update all relevant `ORACLE_BASE`, OLR, inventory, network, and home-property configuration files. Restore permissions with `roothas.sh -init`, move AHF data with `ahfctl movedatadir`, and restart the Oracle Restart stack. All executables and applications using the old base or shared libraries must be stopped beforehand.
Read More: [Changing the Oracle Grid Infrastructure Base Path](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-and-configuring-oracle-grid-infrastructure-for-a-standalone-server/changing-the-oracle-grid-infrastructure-base-path.html)


# Installing Oracle AI Database

## About Installing Oracle AI Database

Oracle AI Database and Oracle Grid Infrastructure are distributed as image-based ZIP files and can be installed through Oracle Universal Installer, silent or response-file methods, or Oracle Fleet Patching and Provisioning. Software is available from Oracle download portals and may also be provided on installation media. Oracle ASM requires a prior Oracle Grid Infrastructure for a standalone server installation, and earlier-release installers cannot install this release.

Read More: [About Installing Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/about-installing-oracle-ai-database.html)

## About Image-Based Oracle AI Database Installation

Image-based installation creates an Oracle home, extracts the database image into it, and runs the setup wizard to register and configure the product. It supports single-instance and clustered installations, upgrades, automation, customized patched images, and enables all database options by default. The image must be extracted directly into an Oracle Optimal Flexible Architecture-compliant Oracle home before starting the wizard.

Read More: [About Image-Based Oracle AI Database Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/about-image-based-oracle-ai-database-installation.html)

## About Deploying Oracle AI Database Using Oracle FPP

Oracle Fleet Patching and Provisioning uses centrally managed gold images of Oracle homes to provision, patch, and upgrade database software. It supports single-instance databases on Oracle Restart, clusters, or standalone nodes, regardless of whether Oracle Grid Infrastructure is installed, across data-center and cloud environments. Rapid Home Provisioning has used the Oracle FPP name since Oracle Database 19c.

Read More: [About Deploying Oracle AI Database Using Oracle FPP](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/about-deploying-oracle-ai-database-using-oracle-fpp.html)

## Downloading Oracle Software

Oracle AI Database software can be downloaded from the Oracle website or Oracle Software Delivery Cloud, subject to the applicable license terms. Downloads should be stored in adequately sized product directories, verified against published file sizes and checksums, and extracted after download. Software Delivery Cloud requires account sign-in, product and platform selection, license acceptance, and digest verification.

Read More: [Downloading Oracle Software](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/downloading-oracle-software.html)

## About Character Set Selection During Installation

The database character set should be selected carefully before creation because changing it later can require extensive data conversion. Oracle recommends the Unicode AL32UTF8 default, which supports broad language and technology compatibility and allows PDBs using most supported character sets to plug into the container database. Non-AL32UTF8 container databases generally permit only PDBs with the same character set; legacy-character-set PDBs can instead be created temporarily and plugged into an AL32UTF8 container.

Read More: [About Character Set Selection During Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/about-character-set-selection-during-installation.html)

## About Automatic Memory Management Installation Options

Typical installations generally enable automatic memory management, while Advanced installations allow manual allocation or automatic management. Automatic memory management assigns a target and dynamically balances memory between the SGA and instance PGA, but it cannot be selected when database instance memory exceeds 4 GB; automatic shared memory management should be used instead. The feature can also be enabled after installation, which requires a database restart.

Read More: [About Automatic Memory Management Installation Options](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/about-automatic-memory-management-installation-options.html)

## Running the Installer in a Different Language

The installer language is determined by the operating system locale. Supported languages include Brazilian Portuguese, French, German, Italian, Japanese, Korean, Simplified Chinese, Spanish, and Traditional Chinese. If the session locale is unsupported, the installer runs in English.

Read More: [Running the Installer in a Different Language](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/running-the-installer-in-a-different-language.html)

## Installing the Oracle AI Database Software

Oracle AI Database can be installed through the graphical setup wizard or command-line and response-file methods, with Oracle Grid Infrastructure required beforehand for Oracle Restart or ASM use. Image-based installations extract `db_home.zip` into an OFA-compliant, non-symlinked Oracle home and start `runInstaller` from that directory; Release Updates can be applied with `-applyRU`, and root scripts can be automated or run manually. The setup supports database creation, gold-image generation, and command-line configuration, and non-CDB databases are no longer supported; EM Express is desupported in Oracle AI Database 26ai.

Read More: [Installing the Oracle AI Database Software](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/installing-the-oracle-ai-database-software.html)

## Installing Oracle AI Database Using RPM Packages

RPM-based installation provides automated prerequisite checks, software extraction, ownership and inventory management, root configuration, and Oracle home creation for supported single-instance Enterprise Edition deployments. Software can be installed through `yum` or manually after installing the preinstallation RPM, then configured with DBCA or the provided service script, which creates a default container database, pluggable database, and listener. RPM patching and RPM-based upgrades are unsupported, Standard Edition 2 is unavailable, and multiple software versions may coexist in separate homes.

Read More: [Installing Oracle AI Database Using RPM Packages](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/installing-oracle-ai-database-using-rpm-packages.html)

## Single-Server Rolling Database Maintenance

Oracle AI Database 26ai supports rolling patches and local maintenance for Oracle RAC or RAC One Node databases on a single-node Grid Infrastructure cluster. A conventional single-instance database must first be converted to RAC or RAC One Node, after which a new database home and instance can be started and the database transferred with minimal interruption. Readiness can be checked with `cluvfy`, and the process can be orchestrated manually or through Oracle Fleet Patching and Provisioning.

Read More: [Single-Server Rolling Database Maintenance](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/installing-oracle-ai-database/single-server-rolling-database-maintenance.html)


# Configuring Oracle Homes

## About Read-Only Oracle Homes

Read-only Oracle Homes separate Oracle software installation from configuration and operational data. Although the Oracle Home is configured as read-only, its file system and mount point remain read/write. Configuration files and logs are relocated outside ORACLE\_HOME into ORACLE\_BASE\_HOME and ORACLE\_BASE\_CONFIG, and read-only mode should be enabled with roohctl before creating the database or listener.

Read More: [About Read-Only Oracle Homes](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/about-read-only-oracle-homes.html)

## About Oracle Base Home

ORACLE\_BASE\_HOME stores user-specific, instance-specific, networking, and log files for both read-only and read/write Oracle Homes. In read/write mode it is the same as ORACLE\_HOME; in read-only mode it is located at ORACLE\_BASE/homes/HOME\_NAME. The orabasehome command, run from ORACLE\_HOME/bin, reports the applicable path.

Read More: [About Oracle Base Home](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/about-oracle-base-home.html)

## About Oracle Base Config

ORACLE\_BASE\_CONFIG stores Oracle Home configuration files, including database files in its dbs subdirectory. In read/write mode it equals ORACLE\_HOME, while in read-only mode it equals ORACLE\_BASE, allowing configuration files such as SPFILEs and password files to reside outside the software installation. The orabaseconfig command reports the applicable location.

Read More: [About Oracle Base Config](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/about-oracle-base-config.html)

## About orabasetab

The orabasetab file, located at ORACLE\_HOME/install/orabasetab, defines ORACLE\_HOME, ORACLE\_BASE, HOME\_NAME, and whether the home is read-only. Its final colon-delimited entry identifies the Oracle Home, base directory, internal home name, and mode flag, with Y indicating read-only and N indicating read/write. In read-only mode, ORACLE\_BASE\_HOME is ORACLE\_BASE/homes/HOME\_NAME and ORACLE\_BASE\_CONFIG is ORACLE\_BASE; in read/write mode, all three logical locations equal ORACLE\_HOME.

Read More: [About orabasetab](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/about-orabasetab.html)

## Location of Networking Files in a Read-Only Oracle Home

In a read-only Oracle Home, networking directories are stored under ORACLE\_BASE\_HOME/network, while read/write installations use ORACLE\_HOME/network. tnsnames.ora is searched for in TNS\_ADMIN, /etc, ORACLE\_BASE\_HOME/network/admin, and ORACLE\_HOME/network/admin; sqlnet.ora is searched for in TNS\_ADMIN, ORACLE\_BASE\_HOME/network/admin, and ORACLE\_HOME/network/admin. Password files are searched for first in ORACLE\_BASE\_CONFIG/dbs and then ORACLE\_HOME/dbs.

Read More: [Location of Networking Files in a Read-Only Oracle Home](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/location-of-networking-files-in-a-read-only-oracle-home.html)

## Enabling a Read-Only Oracle Home

Enable read-only mode after a software-only Oracle AI Database installation by running roohctl -enable, then create the database with DBCA. The Oracle Home must be configured before creating the database or listener. For Oracle RAC, run roohctl on every node or use its nodelist option to specify all cluster nodes.

Read More: [Enabling a Read-Only Oracle Home](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/enabling-a-read-only-oracle-home.html)

## Copying demo Directories to Oracle Base Home

Because demo directories may be modified when used, copy the required directories from ORACLE\_HOME to ORACLE\_BASE\_HOME in a read-only installation. The affected locations include jdbc, odbc, ord/http, precomp, rdbms, sqlplus, and xdk demos; symbolic links must replace the original odbc, precomp, rdbms, and xdk paths as specified. Also copy or synchronize init.ora into ORACLE\_BASE\_HOME. The plsql/demo ncmpdemo.sql file is not usable in read-only mode.

Read More: [Copying demo Directories to Oracle Base Home](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/copying-demo-directories-to-oracle-base-home.html)

## Determining If an Oracle Home is Read-Only

Run the orabasehome command from ORACLE\_HOME/bin to determine the Oracle Home mode. If it returns the same path as ORACLE\_HOME, the installation is read/write; if it returns ORACLE\_BASE/homes/HOME\_NAME, the installation is read-only. The ORACLE\_HOME environment variable must be set before running the command.

Read More: [Determining If an Oracle Home is Read-Only](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/determining-if-an-oracle-home-is-read-only.html)

## File Path and Directory Changes in Read-Only Oracle Homes

Read-only Oracle Homes relocate writable configuration and operational directories outside the software installation. ORACLE\_BASE\_HOME becomes ORACLE\_BASE/homes/HOME\_NAME, ORACLE\_BASE\_CONFIG becomes ORACLE\_BASE, networking files move under ORACLE\_BASE\_HOME/network, database configuration files under ORACLE\_BASE\_CONFIG/dbs, and hs, mgw, and drdaas administration files under ORACLE\_BASE\_HOME. In read/write mode, these locations remain within ORACLE\_HOME; the orabasehome and orabaseconfig commands display the active paths.

Read More: [File Path and Directory Changes in Read-Only Oracle Homes](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/configuring-oracle-homes/file-path-and-directory-changes-in-read-only-oracle-homes.html)


# Oracle AI Database Postinstallation Tasks

Complete configuration tasks after you install Oracle AI Database.

You are required to complete some configuration tasks after Oracle AI Database is installed. In addition, Oracle recommends that you complete additional tasks immediately after installation. You must also complete product-specific configuration tasks before you use those products.

This chapter describes basic configuration only. Refer to product-specific administration and tuning guides for more detailed configuration and tuning information.

## Downloading Release Update Patches

Download and install Oracle Release Updates (RUs) and Monthly Recommended Patches (MRPs) after installation by searching My Oracle Support by patch, bug, product, release, platform, or language criteria. Review each patch’s README before applying it. For Oracle AI Database 26ai, Oracle recommends out-of-place patching with gold images; OPatch and OPatchAuto remain available for in-place patching, while RUs are not supported for the Free release.

Read More: [Downloading Release Update Patches](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/downloading-release-update-patches.html)

## Oracle AI Database Patching

Oracle AI Database 26ai Release Updates are delivered as gold images for out-of-place patching or binary patches for in-place application with OPatch or OPatchAuto. Out-of-place patching involves downloading and extracting the gold image, creating a new Oracle home with `runInstaller -setupDBHomeAs`, and moving the database with DBCA or Oracle FPP Local Mode; in clustered environments, FPP commands must be run from one node and retried from that same node if they fail.

Read More: [Oracle AI Database Patching](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/oracle-ai-database-patching.html)

## Unlocking Accounts and Resetting Oracle AI Database User Passwords

Most administrative accounts are locked or have revoked passwords after installation, so required accounts must be explicitly unlocked and assigned secure passwords. Review predefined and Oracle-maintained accounts carefully, avoid modifying accounts marked as Oracle maintained, and use SQL\*Plus as SYSDBA with `ALTER USER` to lock, unlock, or reset accounts. EM Express is desupported in Oracle AI Database 26ai; use supported database management tools instead.

Read More: [Unlocking Accounts and Resetting Oracle AI Database User Passwords](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/unlocking-accounts-and-resetting-oracle-ai-database-user-passwords.html)

## About Installing Oracle Autonomous Health Framework

Oracle Autonomous Health Framework is preinstalled with Oracle AI Database and provides proactive health checks and diagnostic collection across database, Grid Infrastructure, RAC, GoldenGate, Exadata, SuperCluster, Recovery Appliance, and related deployments. Updates are included with database Release Updates or available through My Oracle Support, and Oracle recommends maintaining the latest version.

Read More: [About Installing Oracle Autonomous Health Framework](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/about-installing-oracle-autonomous-health-framework.html)

## Enabling and Disabling Oracle AI Database Options After Installation

Use the postinstallation `chopt` utility in `ORACLE_HOME/bin` to enable or disable supported database options such as Partitioning and Real Application Testing. The database must be stopped before running the command and restarted afterward; cloned or image-based installations enable all options by default. Oracle Advanced Analytics cannot be disabled with `chopt`, and OLAP-related technologies are deprecated.

Read More: [Enabling and Disabling Oracle AI Database Options After Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/enabling-and-disabling-oracle-ai-database-options-after-installation.html)

## Creating a Fast Recovery Area

Create a fast recovery area to centralize RMAN backups, archived logs, control-file backups, and database copies, enabling faster recovery and automatic obsolete-file management. Configure it with `DB_RECOVERY_FILE_DEST` and `DB_RECOVERY_FILE_DEST_SIZE`, preferably on a separate ASM disk group sized for several days of recovery data or more. Use ASMCA to create and configure the disk group, selecting an appropriate name, redundancy level, and eligible disks.

Read More: [Creating a Fast Recovery Area](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/creating-a-fast-recovery-area.html)

## About Changes in Default SGA Permissions for Oracle AI Database

Oracle AI Database restricts read and write access to the System Global Area to the installation owner by default, improving security but potentially limiting OSDBA administrators. The `ALLOW_GROUP_ACCESS_TO_SGA` parameter controls whether OSDBA group members can access the SGA; its default `FALSE` setting is recommended, although it can be changed to `TRUE` when group read access is required.

Read More: [About Changes in Default SGA Permissions for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/about-changes-in-default-sga-permissions-for-oracle-ai-database.html)

## Setting Language and Locale Preferences for Client Connections

Configure client locale and character-set settings according to the application, operating system, and I/O device. OCI applications use `NLS_LANG` and related settings, with the character-set component matching the device encoding; an incorrect value can cause data loss. JDBC applications instead use the Java VM locale and Unicode internally, while three-tier deployments may define locale settings in application-server configuration.

Read More: [Setting Language and Locale Preferences for Client Connections](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/setting-language-and-locale-preferences-for-client-connections.html)

## Recompile Invalid Objects in the Database

After installation, patching, or upgrading, recompile invalid objects across the CDB and PDBs using Oracle scripts such as `utlrp.sql`, `utlprp.sql`, or `utlprpom.sql`, executed through `catcon.pl`. AutoUpgrade generally recompiles Oracle-maintained objects automatically, while administrators can choose serial or parallel recompilation, process Oracle-owned objects first, and control PDB concurrency and logging through command parameters.

Read More: [Recompile Invalid Objects in the Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/recompile-invalid-objects-in-the-database.html)

## About the oradism Utility in Oracle AI Database

The `oradism` utility enables Oracle AI Database to perform operations requiring privileged operating-system access on Linux. Oracle installs and configures it automatically as a root-owned binary in `$ORACLE_HOME/bin`; its background process, `ora_dism_<sid>`, starts with the database and stops during shutdown. Alert-log errors should prompt verification of the binary’s location and ownership.

Read More: [About the oradism Utility in Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/about-the-oradism-utility-in-oracle-ai-database.html)

## Checking Installed Oracle AI Database Contents and Directory Location

To inspect installed Oracle AI Database products and their locations, launch Oracle Universal Installer from `$ORACLE_HOME/oui/bin`, open Installed Products, select the relevant product, and view its details. Close the inventory and exit the installer after reviewing the installation contents.

Read More: [Checking Installed Oracle AI Database Contents and Directory Location](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/checking-installed-oracle-ai-database-contents-and-directory-location.html)

## Setting the Oracle User Environment Variables

Configure the installation owner’s profile on every cluster node with `ORACLE_BASE`, `ORACLE_HOME`, and `ORACLE_SID`, and add `$ORACLE_HOME/bin` to `PATH`. Correctly setting these variables is required for SQL\*Plus and other Oracle utilities to locate the database installation and instance.
Read More: [Setting the Oracle User Environment Variables](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-postinstallation-tasks/setting-the-oracle-user-environment-variables.html)


# Cloning Oracle AI Database

## Cloning an Oracle Home

Cloning an Oracle home creates a configured copy for a new environment, reducing installation time when deploying multiple Oracle AI Database installations. It is particularly useful for preserving patches already applied to the source home, ensuring the cloned home includes the same updates.

Read More: [Cloning an Oracle Home](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/cloning-oracle-ai-database/cloning-an-oracle-home.html)

## Cloning an Oracle AI Database Home Using an Image File

To clone an Oracle AI Database home with an image, log in as the Oracle installation owner, create a gold image with runInstaller and the -createGoldImage option, extract it into a new OFA-compliant Oracle home, and run the installer there using the software-only configuration. Complete the setup and root-script requirements, then use Oracle Database Configuration Assistant to create a database and configure its listener; Oracle home and base paths, including their parents, must not be symbolic links.

Read More: [Cloning an Oracle AI Database Home Using an Image File](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/cloning-oracle-ai-database/cloning-an-oracle-ai-database-home-using-an-image-file.html)

## Cloning an Oracle AI Database Home Using clone.pl

The clone.pl method is deprecated from Oracle Database 19c onward, so software-only installation is recommended instead. If used, verify the source installation and patches, stop Oracle processes, archive and transfer only the Oracle home, extract it on the target, correct ownership, remove source-specific network and database configuration files, and run clone.pl with the target paths and operating system groups. After cloning, use Net Configuration Assistant and Database Configuration Assistant to configure networking and create a database, while providing root privileges when prompted and reviewing the cloning logs for results.
Read More: [Cloning an Oracle AI Database Home Using clone.pl](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/cloning-oracle-ai-database/cloning-an-oracle-ai-database-home-using-clone.pl.html)


# Removing Oracle AI Database Software

## About Oracle Deinstallation Options

The `deinstall` command, located in each Oracle home’s `$ORACLE_HOME/deinstall` directory, stops and removes supported Oracle software and its configuration, including Oracle AI Database, Grid Infrastructure, Oracle RAC, and Database Client; individual components cannot be removed independently. It can run interactively or use response files generated with `-checkonly`, must match the software release being removed, and requires complete configuration details for inactive or failed installations. Grid Infrastructure removal may require `rootcrs.sh` or `roothas.sh`, ACFS and ADVM preparation, and GNS DNS cleanup. Deinstallation can also delete Oracle base contents such as administrative, diagnostic, recovery, and data directories, including user data and files outside the Oracle base, so Oracle recommends OFA layouts and reserving Oracle paths exclusively for Oracle software.

Read More: [About Oracle Deinstallation Options](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/about-oracle-deinstallation-options.html)

## Oracle deinstall Command

The `deinstall` command removes Oracle software and operating-system configuration for a specific Oracle home. Its options support silent execution (`-silent`), configuration discovery and response-file generation (`-checkonly`), custom response files and parameter overrides (`-paramfile`, `-params`), alternate response, temporary, and log directories (`-o`, `-tmpdir`, `-logdir`), multinode local-node deinstallation (`-local`), retaining local or remote Grid homes while deconfiguring them (`-skipLocalHomeDeletion`, `-skipRemoteHomeDeletion`), and command help (`-help`). Silent mode requires a functioning installation or a suitable response file.

Read More: [Oracle deinstall Command](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/oracle-deinstall-command.html)

## Deinstallation Examples for Oracle AI Database

Run `deinstall` from the target Oracle home’s `deinstall` directory to begin removal without specifying the home path. Use `-checkonly` to inspect the installation and generate a response file, then provide that file with `-paramfile` for a subsequent deinstallation. The same process applies to Oracle Grid Infrastructure by running the command from its Grid home, with response files identifying the corresponding database or Grid installation.

Read More: [Deinstallation Examples for Oracle AI Database](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/deinstallation-examples-for-oracle-ai-database.html)

## Downgrading Oracle Restart

Downgrading Oracle Restart involves backing up the SPFILE, recording database and Oracle Restart resource configurations, downgrading the database when necessary, stopping databases and listeners, and deinstalling Oracle Restart while retaining existing ASM disk groups. A lower-version Grid home is then installed and registered, Oracle Restart is configured, ASM is restored using a temporary disk group and the existing disk group, and the temporary group is removed. Databases, listeners, and services are re-added with their recorded attributes, started, and verified; pluggable databases must be opened and the final resource status checked.

Read More: [Downgrading Oracle Restart](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/downgrading-oracle-restart.html)

## Removing an RPM-Based Oracle AI Database Installation

To remove an RPM-based Oracle AI Database installation, run `yum -y remove oracle-ai-database-ee-26ai` as `root`. The command automatically removes default components such as database `ORCLCDB` and listener `LISTENER`, but stops if additional databases or listeners are configured. In that case, use DBCA as the Oracle software owner to remove databases and NETCA to remove listeners, then rerun the `yum` removal command as `root`.

Read More: [Removing an RPM-Based Oracle AI Database Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/removing-an-rpm-based-oracle-ai-database-installation.html)

## Deinstalling Previous Release Grid Home

To remove a previous-release Grid home after an upgrade, log in as `root`, adjust the home and parent-directory permissions and ownership so the Grid software owner can access them, then switch to the Grid Infrastructure owner and run the release-matching `deinstall` command. The procedure assumes the specified paths, owner, and inventory group are adapted to the local installation.

Read More: [Deinstalling Previous Release Grid Home](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/deinstalling-previous-release-grid-home.html)

## Changing the Oracle Restart Home Path

To change the Oracle Restart Grid home path, first shut down executables and applications using the old home or its shared libraries. Copy the Grid Infrastructure binaries to the new location as `root`, assign ownership to the Grid software owner, run `gridSetup.sh` in software-only mode to relink binaries and update the inventory without changing the Oracle base, and execute the new home’s `root.sh`. Finally, switch to the new home interactively using the patch and zero-downtime switching options or silently with `-switchGridHome`, optionally controlling ACFS driver updates.
Read More: [Changing the Oracle Restart Home Path](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/removing-oracle-ai-database-software/changing-the-oracle-restart-home-path.html)


# Oracle Grid Infrastructure Command-Line Options

Use this comprehensive list of Oracle Grid Infrastructure silent mode commands to install, configure, patch, upgrade, downgrade, and manage Oracle Grid Infrastructure deployments.

## gridSetup.sh Command

The `gridSetup.sh` command launches the Oracle Grid Infrastructure installer when used without options and supports installation, prerequisite checking, configuration-tool execution, debugging, patch application, response files, and automated root-script execution. Silent installations can use response files, while root execution can be configured for direct root or sudo-based operation; the page also provides an example of silent installation with prerequisite failures ignored.

Read More: [gridSetup.sh Command](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-command.html)

## gridSetup.sh -setupHome

The `gridSetup.sh -setupHome` command installs Oracle Grid Infrastructure software and registers it with the central inventory. It requires OSDBA, OSASM, and Oracle base values, and optionally accepts OSOPER, cluster-node, inventory-location, prerequisite, and automated root-script options for deploying the software across a cluster.

Read More: [gridSetup.sh -setupHome](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-setuphome.html)

## gridSetup.sh -setupHomeAs

The `gridSetup.sh -setupHomeAs` command installs Oracle Grid Infrastructure using the Oracle base and privileged operating-system groups from a specified source Grid home, with optional overrides. It supports prerequisite checks, cluster-node selection, Oracle base and group definitions, and automated root-script execution for reproducing a Grid home configuration across nodes.

Read More: [gridSetup.sh -setupHomeAs](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-setuphomeas.html)

## gridSetup.sh -configureStandaloneServer

The `gridSetup.sh -configureStandaloneServer` command configures Oracle Grid Infrastructure for a standalone server, also known as Oracle Restart. It defines required operating-system groups, Oracle base, ASM disk-group name, and storage either through a disk list or failure-group configuration, while supporting ASM redundancy and allocation settings, disk discovery, Enterprise Manager integration, inventory registration, prerequisite checks, configuration tools, and root-script execution.

Read More: [gridSetup.sh -configureStandaloneServer](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-configurestandaloneserver.html)

## gridSetup.sh -switchGridhome

The `gridSetup.sh -switchGridHome` command changes the active Grid Infrastructure home to the current home after software-only installation has been completed on all cluster nodes. It supports zero-downtime switching, optional driver-update suppression, rollback to an older home, prerequisite checks, service-drain timeouts, and resuming pending or failed jobs; zero-downtime patching and rollback cannot be combined.

Read More: [gridSetup.sh -switchGridhome](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-switchgridhome.html)

## gridSetup.sh -createGoldImage

The `gridSetup.sh -createGoldImage` command creates a reusable gold image from the current Oracle Grid Infrastructure home. It requires a destination directory and optionally accepts a comma-separated list of absolute files or directories to exclude, including wildcarded directory paths that exclude entire directory trees.

Read More: [gridSetup.sh -createGoldImage](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-creategoldimage.html)

## gridSetup.sh -upgrade

The `gridSetup.sh -upgrade` command upgrades an earlier Oracle Grid Infrastructure installation to the current release. It supports prerequisite validation, Oracle base specification, readiness-only dry runs, cluster batch mapping, ignoring unavailable nodes, configuration-tool execution, and automated root-script execution.

Read More: [gridSetup.sh -upgrade](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-upgrade.html)

## gridSetup.sh -downgrade

The `gridSetup.sh -downgrade` command restores Oracle Grid Infrastructure to an earlier release following a failed upgrade. It requires the path to the previous Grid home and optionally supports configuration-tool execution and automated root-script execution.

Read More: [gridSetup.sh -downgrade](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-grid-infrastructure-command-line-options/gridsetup.sh-downgrade.html)


# Oracle AI Database Installer Command-Line Options

Use this comprehensive list of Oracle AI Database installer silent mode commands to install, configure, and manage Oracle AI Database.

## runInstaller Command

The `runInstaller` command supports general installer operations through optional flags, including silent or response-file installations, prerequisite validation, debugging, one-off patch application, configuration-tool execution, and automatic root-script execution using either root or sudo. Options also control inventory locations, sudo credentials, stack checks, and help output; for example, a silent installation can use a response file while ignoring prerequisite failures.

Read More: [runInstaller Command](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installer-command-line-options/runinstaller-command.html)

## runInstaller -setupDBHome

The `runInstaller -setupDBHome` command installs Oracle AI Database software and registers it with the central inventory. It requires the Oracle base directory and operating-system groups for SYSDBA, SYSBACKUP, SYSDG, SYSKM, and SYSRAC, while optionally supporting SYSOPER, prerequisite checks, automatic root-script execution, cluster-node lists, Enterprise or Standard Edition 2 selection, and a specified inventory location.

Read More: [runInstaller -setupDBHome](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installer-command-line-options/runinstaller-setupdbhome.html)

## runInstaller -setupDBHomeAs

The `runInstaller -setupDBHomeAs` command installs Oracle AI Database software using the Oracle base and privileged operating-system group configuration from a specified source Oracle home. Optional parameters allow overriding group assignments, Oracle base, cluster nodes, database edition, prerequisite execution, and root-script execution, with support for multi-node installations modeled on the source home.

Read More: [runInstaller -setupDBHomeAs](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installer-command-line-options/runinstaller-setupdbhomeas.html)

## runInstaller -createDatabase

The `runInstaller -createDatabase` command installs Oracle AI Database software and creates a single-instance database, including a container database and pluggable database, when no existing Oracle home is available. It requires privileged operating-system groups, Oracle base, database identity values, memory allocation, and either a filesystem data location or ASM disk group; optional features include recovery, automatic memory management, database edition and type selection, Enterprise Manager integration, shared schema passwords, configuration tools, inventory specification, and automatic root-script execution.

Read More: [runInstaller -createDatabase](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installer-command-line-options/runinstaller-createdatabase.html)

## runInstaller -createGoldImage

The `runInstaller -createGoldImage` command creates a gold image from the current Oracle AI Database home and stores it in a required destination directory. It runs silently and optionally accepts a comma-separated list of absolute files or directories to exclude, including wildcard-based directory trees, allowing specific files, patch storage, or multiple paths to be omitted from the image.
Read More: [runInstaller -createGoldImage](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/oracle-ai-database-installer-command-line-options/runinstaller-creategoldimage.html)


# Completing Preinstallation Tasks Manually

You can complete the preinstallation configuration tasks manually.

Oracle recommends that you use Oracle Universal Installer and Configuration Verification Utility (CVU) fixup scripts to complete minimal configuration settings. If you cannot use fixup scripts, then complete minimum system settings manually.

## Configuring Kernel Parameters for Linux

This page explains how to manually configure Linux kernel parameters when fixup scripts or the Oracle AI Database Preinstallation RPM are unavailable or insufficient. It lists minimum semaphore, shared-memory, file, asynchronous I/O, network port, and socket buffer settings; provides commands to inspect current values; and describes persisting changes in `/etc/sysctl.d/97-oracle-database-sysctl.conf` before applying and verifying them with `sysctl`. It also covers configuration guidelines, SUSE-specific `hugetlb_shm_group` settings, and manual UDP/TCP ephemeral-port configuration, recommending a range beginning at 9000, appropriate even/odd endpoints, sufficient capacity, and persistent application after restart.

Read More: [Configuring Kernel Parameters for Linux](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/completing-preinstallation-tasks-manually/configuring-kernel-parameters-for-linux.html)

## Configuring Default Thread Limits Value for SUSE Linux

This page instructs administrators using SUSE Linux Enterprise Server 15 or later to increase the systemd `DefaultTasksMax` limit from its typical value of 512 to 65535 to prevent service failures caused by thread limits. Administrators should inspect the current setting with `systemctl show --property DefaultTasksMax` or `/etc/systemd/system.conf`, uncomment and set the configuration value if necessary, then activate it by rebooting or running `systemctl daemon-reload`.

Read More: [Configuring Default Thread Limits Value for SUSE Linux](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/completing-preinstallation-tasks-manually/configuring-default-thread-limits-value-for-suse-linux.html)

## Configuring Storage Device Path Persistence Manually

This page describes manually preserving Oracle ASM storage device names, ownership, and permissions across Linux restarts by creating custom udev rules. It explains collecting stable SCSI identifiers with `scsi_id`, enabling trusted-device handling through `/etc/scsi_id.config`, and creating rules that assign the installation owner, ASM OSDBA group, and mode 0660 in the required attribute order. On clusters, the rules file must be copied to every node; administrators should run `partprobe`, test the rules with `udevtest`, reload or restart udev according to the Linux distribution, and verify the resulting device ownership and permissions.
Read More: [Configuring Storage Device Path Persistence Manually](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/completing-preinstallation-tasks-manually/configuring-storage-device-path-persistence-manually.html)


# Response Files

## Reasons for Using Silent Mode or Response File Mode

Silent mode supports unattended, scheduled, repeatable installations across multiple systems and installations on systems without X Window System software; it shows progress in the terminal without displaying installer screens. Response file mode supports similar installations while supplying default answers to some or all installer prompts.

Read More: [Reasons for Using Silent Mode or Response File Mode](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/response-files/reasons-for-using-silent-mode-or-response-file-mode.html)

## Using Response Files

Response files enable Oracle products to be installed and configured in silent or response file mode. Complete all prerequisite tasks first, prepare and run the response file, execute root scripts as prompted, and for software-only installations run NETCA and Oracle DBCA in the appropriate noninteractive mode to configure the listener and database.

Read More: [Using Response Files](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/response-files/using-response-files.html)

## Preparing Response Files

Oracle provides product-specific response file templates in the Oracle home or Grid home response directories, including templates for database installation, DBCA, NETCA, and Grid Infrastructure setup. Copy and edit the appropriate template, retain the `.rsp` suffix, validate all required values, and secure the file with ownership restricted to the software owner and permissions set to 600 because it may contain plain-text passwords. Response files can also be recorded from OUI’s interactive Summary screen; recorded files exclude passwords and should be reviewed and updated before reuse on another system.

Read More: [Preparing Response Files](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/response-files/preparing-response-files.html)

## Running Oracle Universal Installer Using a Response File

Run `runInstaller` for Oracle AI Database or `gridSetup.sh` for Grid Infrastructure with a fully qualified response file path, optionally adding `-silent`; use `-help` to view available options. Complete normal prerequisites, run as the software owner, set `DISPLAY` only for response file mode, and execute `orainstRoot.sh` and `root.sh` as root when prompted. Relative response file paths are not supported.

Read More: [Running Oracle Universal Installer Using a Response File](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/response-files/running-oracle-universal-installer-using-a-response-file.html)

## Postinstallation Configuration Using Response File Created During Installation

An installation response file can be reused with `-executeConfigTools` to run postinstallation configuration assistants for Oracle AI Database or Grid Infrastructure. Before execution, add the required administrative passwords, secure the file with 600 permissions and appropriate ownership, then run the product configuration command from the relevant Oracle home; add `-silent` when graphical progress display is not desired. Password requirements depend on enabled features such as ASM, IPMI, and Enterprise Manager.

Read More: [Postinstallation Configuration Using Response File Created During Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/response-files/postinstallation-configuration-using-response-file-created-during-installation.html)

## Running Configuration Assistants Using Response Files

Configuration assistants can run noninteractively by copying and editing their templates, including `dbca.rsp` for Oracle DBCA and `netca.rsp` for NETCA. DBCA uses `-responseFile` with `-createDatabase`, requires `DISPLAY`, and may include `-silent`; NETCA uses `-silent -responsefile` to configure listeners, naming methods, and service names. Correct response file values, the Oracle home environment, and full file paths are required, or the assistants fail.
Read More: [Running Configuration Assistants Using Response Files](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/response-files/running-configuration-assistants-using-response-files.html)


# Optimal Flexible Architecture

## About the Optimal Flexible Architecture Standard

Oracle Optimal Flexible Architecture (OFA) provides directory, naming, permission, and configuration guidelines for organizing Oracle software and databases. It supports coexistence of multiple database versions and owners, simplifies administration and expansion, and helps associate Oracle Base locations with Automatic Diagnostic Repository data for incident collection. Oracle recommends accepting the installer’s OFA defaults, particularly for large or multi-database environments.

Read More: [About the Optimal Flexible Architecture Standard](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/optimal-flexible-architecture/about-the-optimal-flexible-architecture-standard.html)

## About Multiple Oracle Homes Support

Oracle AI Database supports multiple Oracle homes on the same system, allowing different releases and databases to coexist when each release is installed in a separate home. OFA-compliant naming and mount points simplify administration, protect user home directories, and support multiple owners and versions. Each installation requires extracting the image and running setup from its own home; newer releases can access older databases only for upgrades, and multiple releases must not share one Oracle home.

Read More: [About Multiple Oracle Homes Support](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/optimal-flexible-architecture/about-multiple-oracle-homes-support.html)

## About the Oracle Inventory Directory and Installation

The Oracle Inventory directory, or oraInventory, records all Oracle software installations and is shared by installation owners through the OINSTALL group. Existing installations use the inventory identified by `/etc/oraInst.loc`; new installations determine its location from ORACLE\_BASE, an available OFA path, or the installing user’s home directory. Oracle Universal Installer places database homes beneath Oracle Base, while Grid Infrastructure homes use distinct locations and may become root-owned after installation.

Read More: [About the Oracle Inventory Directory and Installation](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/optimal-flexible-architecture/about-the-oracle-inventory-directory-and-installation.html)

## Oracle Base Directory Naming Convention

Oracle Base directories should follow the `/pm/h/u` pattern, combining a mount point, standard directory, and installation owner name. A base can support multiple installations, but separate operating system owners require separate bases, preferably with unique owners for log isolation. Consistent mount paths simplify inventory management, and neither Oracle Base nor its parent directories may be symbolic links.

Read More: [Oracle Base Directory Naming Convention](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/optimal-flexible-architecture/oracle-base-directory-naming-convention.html)

## Oracle Home Directory Naming Convention

Oracle homes conventionally use the `/pm/s/u/product/v/type_n` structure, identifying the mount point, standard directory, owner, software version, installation type, and optional installation counter. A typical database home is `/u01/app/oracle/product/23.0.0/dbhome_1`; separate counters permit multiple products of the same type under one Oracle Base. Oracle homes and all parent directories up to root must be physical directories rather than symbolic links.

Read More: [Oracle Home Directory Naming Convention](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/optimal-flexible-architecture/oracle-home-directory-naming-convention.html)

## Optimal Flexible Architecture File Path Examples

OFA-compliant layouts organize inventory, Oracle Base and home directories, administration files, recovery areas, data files, client software, and Grid Infrastructure across structured mount points such as `/u01` through `/u04`. They support multiple database homes and databases while distinguishing standalone-server Grid homes from cluster Grid homes, with Grid binaries outside the Grid Base for clusters. Oracle ASM is recommended for redundancy and throughput, and private or application files should not be stored within OFA directories.
Read More: [Optimal Flexible Architecture File Path Examples](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/optimal-flexible-architecture/optimal-flexible-architecture-file-path-examples.html)


# Managing Oracle AI Database Port Numbers

## About Managing Ports

Oracle AI Database assigns component and service ports during installation from predefined ranges, selecting the lowest available port after checking for conflicts with other Oracle installations, running processes, and entries in `/etc/services`. Administrators should review assigned ports, prevent duplicate usage, identify active ports with `/bin/netstat -a`, and change assignments after installation when necessary.

Read More: [About Managing Ports](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/managing-oracle-ai-database-port-numbers/about-managing-ports.html)

## Oracle AI Database Component Port Numbers and Protocols

Single-instance installations configure Oracle Net Services Listener on TCP port 1521 by default, selecting the next available port when necessary and allowing manual changes through Net Configuration Assistant. Oracle Connection Manager uses TCP port 1630 and is configured manually in `cman.ora`; Oracle XML DB HTTP and FTP services default to port 0 and require manual port configuration for HTTP or FTP access.
Read More: [Oracle AI Database Component Port Numbers and Protocols](https://docs.oracle.com/en/database/oracle/oracle-database/26/ladbi/managing-oracle-ai-database-port-numbers/oracle-ai-database-component-port-numbers-and-protocols.html)