Oracle Backend for Microservices and AI
Toggle Dark/Light/Auto mode Toggle Dark/Light/Auto mode Toggle Dark/Light/Auto mode Back to homepage
  1. Oracle Backend for Microservices and AI
  2. /
  3. Custom Installations

Custom Installations

The Oracle Backend for Microservices and AI is available to install in your own “custom” environment, which may be an on-premises data center environment, a different cloud provider, or a developer’s desktop.

Prerequisites

You must meet the following prerequisites to use the Oracle Backend for Microservices and AI On-Premises. You need:

  • Access to Oracle Database Enterprise Edition 19.3.0.0
  • Access to a Container Repository
  • Access to a Kubernetes cluster
  • Python 3+

When installing in a desktop environment, for example a developer’s desktop, the previously mentioned pre-requisites may be met through an additional setup task, but there are additional desktop system or software requirements. For example:

  • 2 CPUs or more
  • 8 GB of free memory
  • 60 GB of free disk space (40 GB minikube and container images, 20 GB database)
  • Internet connection
  • Minikube or similar
  • Podman1 or similar
  • Oracle Single Sign-On (SSO) account to download the database image

Download

Download the latest release of Oracle Backend for Microservices and AI.

Setup

A custom installation consists of defining the infrastructure followed by running the Configuration Management Playbook to build images and deploy the Microservices.

For a custom installation, you need to have a Kubernetes cluster and the kubectl command-line interface must be configured to communicate with your cluster.

A Helper Playbook has been provided for desktop installations to assist in defining a specific infrastructure consisting of podman and minikube as outlined in the example documentation:

If your infrastructure does not match that defined in the above examples, do not run the Helper Playbook.

Download the Database or Oracle REST Data Services (ORDS) Images (Desktop Installation)

The desktop installation provisions an Oracle database to the Kubernetes cluster. The images must be downloaded from Oracle’s Container Registry before continuing.

After installing Podman, process these steps:

  1. Log in to Oracle Cloud Infrastructure Registry (Container Registry). For example:

    podman login container-registry.oracle.com

  2. Pull the database image. For example:

    podman pull container-registry.oracle.com/database/enterprise:19.3.0.0

  3. Pull the ORDS image. For example:

    podman pull container-registry.oracle.com/database/ords:21.4.2-gh

Defining the Application

The application is defined in ansible/vars/ebaas.yaml. For example:

---
ebaas_edition: "COMMUNITY"
vault: ""
vault_key: ""
vault_crypto_endpoint: ""
vault_management_endpoint: ""
vault_storage_account_name: "N/A"
vault_storage_account_key: "N/A"
vault_storage: ""
vault_storage_lock: ""
apisix_admin_password: "Correct-horse-Battery-staple-35"
grafana_admin_password: "Correct-horse-Battery-staple-35"
oractl_admin_password: "Correct-horse-Battery-staple-35"
oractl_user_password: "Correct-horse-Battery-staple-35"
...

Create the ansible/vars/ebaas.yaml file, setting values as required. If this is a desktop installation, this file will be created for you by the Desktop Helper playbook.

Defining the Database

The database is defined in ansible/roles/database/vars/main.yaml. For example:

---
database_oracle_dbs: ["BAASPDB"]
database_default_db: BAASPDB
BAASPDB: # noqa: var-naming[pattern]
  username: "PDBADMIN"
  password: "Correct-horse-Battery-staple-35"
  type: "EXTERNAL"
  service: "(DESCRIPTION=(ADDRESS=(PROTOCOL=tcp)(HOST=localhost)(PORT=1521))(CONNECT_DATA=(SERVICE_NAME=BAASPDB)))"
  ocid: ""
...

Create the ansible/roles/database/vars/main.yaml file, setting values as required. If this is a desktop installation, this file will be created for you by the Desktop Helper playbook.

The database_oracle_dbs and database_default_db key values should be the name of your Pluggable Database (PDB). These are followed by the PDB name and Key/Values defining how to access the PDB.

The type can be either:

  • EXTERNAL: A Pre-existing Oracle Database. Define service and leave ocid blank.
  • ADB-S: A Pre-Existing Oracle Autonomous Serverless (ADB-S) database; provide the ocid for the ADB-S. Leave service blank.
  • SIDB_CONTAINER: This will create a 19c containerized database inside the Kubernetes Cluster as part of the deployment. Leave ocid and service blank.

Defining the Container Repository

The Container Repository is defined in ansible/roles/registry/vars/main.yaml. For example:

---
registry_username: "oracle"
registry_password: "Correct-horse-Battery-staple-35"
registry_push_url: "docker.io/myorg"
registry_push_auth:
  auths:
    docker.io/myorg:
      auth: "b3JhY2xlOjdaUVgxLXhhbFR0NTJsS0VITlA0"
registry_pull_url: "docker.io/myorg"
registry_pull_auth:
  auths:
    docker.io/myorg:
      auth: "b3JhY2xlOjdaUVgxLXhhbFR0NTJsS0VITlA0"
...

Create the ansible/roles/registry/vars/main.yaml file, setting values as required. If this is a desktop installation, this file will be created for you by the Desktop Helper playbook.

Specify the URL or authentication credentials for your Container Repository in registry_pull_url, registry_push_url, registry_username, and registry_password.

For the registry_pull_auth and registry_push_auth sections, manually log into your repository and copy the values found in the created file, located in $HOME/.config/containers/auth.json

There may be duplication between the push and pull URL’s. The pull URL is used inside the Pods while the push URL is used from the deployment machine. If you have a private registry inside the Kubernetes cluster, these URL’s could be different. This is the case for the Desktop installation. For example, the push URL is localhost:5000, while the pull URL is <Registry Pod ClusterIP>:5000.

Configuration Management Playbook

From the source package, run the Configuration Management Playbook.

Install Ansible

Using Python, install Ansible to run the Configuration Management Playbook. The Helper script creates a Python virtual environment and installs Ansible along with other additional modules. For example:

./setup_ansible.sh
source ./activate.env

Desktop Helper Playbook

If this is a desktop installation, then a Helper Playbook can be used to help define the infrastructure. Note that this playbook should only be run if following the desktop installation examples:

Run the Desktop Helper Playbook to define the infrastructure. For example:

ansible-playbook ansible/desktop-apply.yaml

Copy kubeconfig

For the desktop installation, this step will have been performed by the Desktop Helper playbook.

Copy the kubeconfig file to ansible/roles/kubernetes/files

Build and Push Images to the Container Repository

For the desktop installation, start a new terminal and tunnel or port-forward to the Minikube cluster. Refer to the specific platform details for more information.

For both installations, run the Images Playbook on the original terminal. For example:

ansible-playbook ansible/images_build.yaml

Install the Microservices

Install the Microservices by running this command:

ansible-playbook ansible/k8s_apply.yaml -t full

Footnotes

  1. Certification has been performed against Podman. However, other container or virtual machine managers are available and may be substituted. ↩︎

gdoc_arrow_left_alt Multi-Cloud (OCI/Azure) Installation MacOS Ventura (x86) gdoc_arrow_right_alt