# Docker Compose
> **Warning**
>
> To provide stability and scalability, we recommend using container orchestration in production.\
> Check the [Kubernetes with Helm](kubernetes.md) example to learn how to run CKEditor AI On-Premises in a cluster.
> **Note**
>
> A valid license key is needed in order to install CKEditor AI On-Premises.\
> [Contact us](https://ckeditor.com/contact/) for a trial license key.
For evaluation and testing purposes, you can follow the [Quickstart checklist](../quickstart-checklist.md), where we provide an easy-to-use way for setting up all necessary infrastructure for CKEditor AI On-Premises on your local machine.
## Setting up the application using the Docker container
Follow the steps below to install CKEditor AI On-Premises using the Docker container.
1. Use the instructions from the [Deployment](../deployment.md#step-1-get-your-download-token) guide to get a download token from the [Customer Portal](https://portal.ckeditor.com/), log into the CKEditor AI On-Premises Docker Registry, and pull the Docker image.
2. Containerize the application using [`docker`](#containerize-example-using-docker) or [`docker-compose`](#containerize-example-using-docker-compose).
3. Create the `Environment` with an `Access Key` via the Management Panel – see [Create an Environment and Access Key](../deployment.md#step-4-create-an-environment-and-access-key). Use this data to [create the token endpoint](../deployment.md#step-5-create-the-token-endpoint).
4. Run the smoke test to verify if the system works properly – see the [Quickstart checklist](../quickstart-checklist.md).
5. Configure CKEditor 5 to communicate with the CKEditor AI On-Premises application – see [Configure CKEditor 5](../deployment.md#step-6-configure-ckeditor-5-to-use-your-ckeditor-ai-on-premises-instance).
## Containerize example using `docker-compose`
Use the steps below to containerize the application using `docker-compose`.
> **Warning**
>
> Please bear in mind that `docker-compose` is useful for development purposes. This, however, does not fulfill our [architecture recommendations](../architecture.md) for production environments.
### Example with MySQL
1. Create the `docker-compose.yml` file:
```yaml
services:
mysql-database:
image: mysql:8.4
environment:
MYSQL_ROOT_PASSWORD: password
volumes:
- ./init.sql:/docker-entrypoint-initdb.d/init.sql:ro
healthcheck: # Define healthcheck to be able to use the `service_healthy` condition.
test: ["CMD-SHELL", "exit | mysql -h localhost -P 3306 -u root -p$$MYSQL_ROOT_PASSWORD"]
interval: 10s
timeout: 30s
retries: 5
redis:
image: redis:7.4
ckeditor-ai:
image: docker.cke-cs.com/ai-service:[version]
depends_on:
redis:
condition: service_started
mysql-database:
condition: service_healthy # Make sure your database is ready for migration.
ports:
- "8000:8000"
restart: always
init: true
environment:
DATABASE_DRIVER: mysql
DATABASE_HOST: mysql-database
DATABASE_USER: root
DATABASE_PASSWORD: password
DATABASE_DATABASE: ai-service-on-premises
REDIS_HOST: redis
STORAGE_DRIVER: database
PROVIDERS: '{"openai":{"type":"openai","apiKeys":["your-api-key"]}}'
ENVIRONMENTS_MANAGEMENT_SECRET_KEY: secret
LICENSE_KEY: your_license_key
```
2. Create the `init.sql` file:
```sql
CREATE DATABASE `ai-service-on-premises`
DEFAULT CHARACTER SET utf8mb4
DEFAULT COLLATE utf8mb4_bin;
```
3. Run:
```bash
docker-compose up
```
### Example with PostgreSQL
1. Create the `docker-compose.yml` file:
```yaml
services:
postgres-database:
image: postgres:17
environment:
- POSTGRES_PASSWORD=password
volumes:
- ./init.sql:/docker-entrypoint-initdb.d/init.sql:ro
healthcheck: # Define healthcheck to be able to use the `service_healthy` condition.
test: pg_isready -U postgres
interval: 10s
timeout: 30s
retries: 5
redis:
image: redis:7.4
ckeditor-ai-postgres:
image: docker.cke-cs.com/ai-service:[version]
depends_on:
redis:
condition: service_started
postgres-database:
condition: service_healthy # Make sure your database is ready for migration.
ports:
- "8000:8000"
restart: always
init: true
environment:
DATABASE_DRIVER: postgres
DATABASE_HOST: postgres-database
DATABASE_PORT: 5432
DATABASE_USER: postgres
DATABASE_PASSWORD: password
DATABASE_DATABASE: cksource
DATABASE_SCHEMA: ai-service-on-premises
REDIS_HOST: redis
STORAGE_DRIVER: database
PROVIDERS: '{"openai":{"type":"openai","apiKeys":["your-api-key"]}}'
ENVIRONMENTS_MANAGEMENT_SECRET_KEY: secret
LICENSE_KEY: your_license_key
```
2. Create the `init.sql` file:
```sql
CREATE DATABASE "cksource";
\connect "cksource";
CREATE SCHEMA "ai-service-on-premises";
```
3. Run:
```bash
docker-compose up
```
### Example with Collaboration Server
CKEditor AI On-Premises can share the SQL database and Redis with Collaboration Server On-Premises 5.0.0 or newer. Both services then use the same environments. See [Collaboration Server integration](../collaboration-server-integration.md) for details.
1. Create the `docker-compose.yml` file:
```yaml
services:
mysql-database:
image: mysql:8.4
environment:
MYSQL_ROOT_PASSWORD: password
volumes:
- ./init.sql:/docker-entrypoint-initdb.d/init.sql:ro
healthcheck: # Define healthcheck to be able to use the `service_healthy` condition.
test: ["CMD-SHELL", "exit | mysql -h localhost -P 3306 -u root -p$$MYSQL_ROOT_PASSWORD"]
interval: 10s
timeout: 30s
retries: 5
redis:
image: redis:7.4
ckeditor-cs:
image: docker.cke-cs.com/cs:[version]
depends_on:
redis:
condition: service_started
mysql-database:
condition: service_healthy # Make sure your database is ready for migration.
ports:
- "8000:8000"
restart: always
init: true
environment:
DATABASE_DRIVER: mysql
DATABASE_HOST: mysql-database
DATABASE_USER: root
DATABASE_PASSWORD: password
DATABASE_DATABASE: cs-on-premises
REDIS_HOST: redis
AI_API_BASE_URL: http://ckeditor-ai:8000 # The address Collaboration Server, including its Management Panel, uses to reach CKEditor AI.
ENVIRONMENTS_MANAGEMENT_SECRET_KEY: secret
LICENSE_KEY: your_collaboration_server_license_key
ckeditor-ai:
image: docker.cke-cs.com/ai-service:[version]
depends_on:
ckeditor-cs:
condition: service_started
ports:
- "8001:8000"
restart: always
init: true
environment:
DATABASE_DRIVER: mysql
DATABASE_HOST: mysql-database
DATABASE_USER: root
DATABASE_PASSWORD: password
DATABASE_DATABASE: cs-on-premises
REDIS_HOST: redis
STORAGE_DRIVER: database
PROVIDERS: '{"openai":{"type":"openai","apiKeys":["your-api-key"]}}'
ENVIRONMENTS_MANAGEMENT_SECRET_KEY: secret
LICENSE_KEY: your_ckeditor_ai_license_key
```
> **Note**
>
> The version of `cs` and `ai-service` images should be the same.\
> You should also connect the services to the same Redis and SQL databases.
2. Create the `init.sql` file:
```sql
CREATE DATABASE `cs-on-premises`
DEFAULT CHARACTER SET utf8mb4
DEFAULT COLLATE utf8mb4_bin;
```
3. Run:
```bash
docker-compose up
```
### Notes
* Use environment variables to configure the application. See the [Required configuration](../configuration.md) guide.
* You should override the `ENVIRONMENTS_MANAGEMENT_SECRET_KEY` variable using a unique and hard to guess string due to security reasons.
* The application will not start without a correct `LICENSE_KEY`.
* The application will not start without at least one LLM provider in `PROVIDERS` – see [LLM providers](../llm-providers.md).
* In the example with Collaboration Server, CKEditor AI On-Premises is served on port `8001`. Set `ai.serviceUrl` in the CKEditor 5 configuration to `http://localhost:8001/v1`.
## Containerize example using Docker
Launch the Docker container:
```bash
docker run --init -p 8000:8000 -e DATABASE_HOST=[your_mysql_host] -e DATABASE_USER=[your_mysql_user] -e DATABASE_PASSWORD=[your_mysql_password] -e DATABASE_DATABASE=[your_database_name] -e REDIS_HOST=[your_redis_host] -e STORAGE_DRIVER=database -e PROVIDERS='{"openai":{"type":"openai","apiKeys":["your-api-key"]}}' -e ENVIRONMENTS_MANAGEMENT_SECRET_KEY=[your_management_secret] -e LICENSE_KEY=[your_license_key] docker.cke-cs.com/ai-service:[version]
```
Notes:
* Use environment variables to configure the application. See the [Required configuration](../configuration.md) guide, or generate the command with the [Setup wizard](../setup-wizard.md). Examples:
* If an SQL database and Redis are hosted on `192.168.118.27`, you should pass the address through the `DATABASE_HOST` and `REDIS_HOST` variables:
```bash
docker run --init -p 8000:8000 -e DATABASE_HOST=192.168.118.27 -e REDIS_HOST=192.168.118.27 -e DATABASE_USER=root -e DATABASE_PASSWORD=root_password -e DATABASE_DATABASE=ai-service-on-premises -e STORAGE_DRIVER=database -e PROVIDERS='{"openai":{"type":"openai","apiKeys":["your-api-key"]}}' -e ENVIRONMENTS_MANAGEMENT_SECRET_KEY=[your_management_secret] -e LICENSE_KEY=[your_license_key] docker.cke-cs.com/ai-service:[version]
```
* If you want to change the port on which the application will be served to `5000`, change the exposed ports:
```bash
docker run --init -p 5000:8000 -e DATABASE_HOST=192.168.118.27 -e REDIS_HOST=192.168.118.27 -e DATABASE_USER=root -e DATABASE_PASSWORD=root_password -e DATABASE_DATABASE=ai-service-on-premises -e STORAGE_DRIVER=database -e PROVIDERS='{"openai":{"type":"openai","apiKeys":["your-api-key"]}}' -e ENVIRONMENTS_MANAGEMENT_SECRET_KEY=[your_management_secret] -e LICENSE_KEY=[your_license_key] docker.cke-cs.com/ai-service:[version]
```
* You should override the `ENVIRONMENTS_MANAGEMENT_SECRET_KEY` variable using a unique and hard to guess string due to security reasons.
* The application will not start without a correct `LICENSE_KEY`.
---
Full index of the Cloud Services documentation: [llms.txt](../../../../llms.txt)