Docker for MindsDB - Mindsdb

MindsDB provides Docker images that facilitate running MindsDB in Docker containers.

<Tip> As MindsDB integrates with numerous [data sources](/integrations/data-overview) and [AI frameworks](/integrations/ai-overview), each integration requires a set of dependencies. Hence, MindsDB provides multiple Docker images for different tasks, as outlined below.

mindsdb/mindsdb:latest (or mindsdb/mindsdb) It is the lightweight Docker image of MindsDB that comes with mysql,postgresql,snowflake,bigquery,mssql and salesforce pre-installed.
mindsdb/mindsdb:huggingface It is the Docker image of MindsDB that comes with the Hugging Face integration preloaded.
</Tip>

Prerequisites

Before proceeding, ensure you have installed Docker, following the official Docker documentation.

Setup

This setup of MindsDB uses one of the available Docker images, as listed above.

<Note> When running MindsDB in one container and the integration you want to connect to it (such as Ollama or PGVector) in another container, then use `http://host.docker.internal` instead of `localhost` when connecting this integration to MindsDB. </Note>

Follow the steps to set up MindsDB in a Docker container.

Install MindsDB

Run this command to create a Docker container with MindsDB:

bash

docker run --name mindsdb_container \
-e MINDSDB_APIS=http,mysql \
-p 47334:47334 -p 47335:47335 \
mindsdb/mindsdb

Where:

docker run is a native Docker command used to spin up a container.
--name mindsdb_container defines a name for the container.
-e MINDSDB_APIS=http,mysql defines the APIs to be exposed by the MindsDB instance. All available APIs include http, mysql, and postgres.
-p 47334:47334 -p 47335:47335 defines the ports where the APIs are exposed (HTTP and MySQL respectively).
mindsdb/mindsdb is a Docker image provided by MindsDB. You can choose a different one from the list above.

Once the container is created, you can use the following commands:

docker stop mindsdb_container to stop the container. Note that this may not always be necessary because when turning off the host machine, the container will also be shut down.
docker start mindsdb_container to restart a stopped container with all its previous changes (such as any dependencies that were installed) intact. Note that docker start restarts a stopped container, while docker run creates a new container.

<Tip> If you don't want to follow the logs and get the prompt back, add the `-d` flag that stands for *detach*.

bash

docker run --name mindsdb_container -e MINDSDB_APIS=http -d -p 47334:47334 mindsdb/mindsdb

</Tip> <Tip> If you want to persist your models and configurations in the host machine, run these commands:

bash

mkdir mdb_data
docker run --name mindsdb_container -e MINDSDB_APIS=http -p 47334:47334 -v $(pwd)/mdb_data:/root/mdb_storage mindsdb/mindsdb

Where -v $(pwd)/mdb_data:/root/mdb_storage maps the newly created folder mdb_data on the host machine to the /root/mdb_storage inside the container. </Tip>

Now you can access the MindsDB editor by going to 127.0.0.1:47334 in your browser.

<p align="center"> </p> <Warning> If you experience any issues related to MKL or your training process does not complete, please add the `MKL_SERVICE_FORCE_INTEL` environment variable, as below.

bash

docker run --name mindsdb_container -e MKL_SERVICE_FORCE_INTEL=1 -e MINDSDB_APIS=http -p 47334:47334 mindsdb/mindsdb

</Warning> <Tip> If you want to enable authentication for MindsDB, you do so by passing the `MINDSDB_USERNAME` and `MINDSDB_PASSWORD` environment variables when running the container.

bash

docker run --name mindsdb_container -e MINDSDB_USERNAME='admin' -e MINDSDB_PASSWORD='password' -e MINDSDB_APIS=http -p 47334:47334 mindsdb/mindsdb

</Tip>

Install dependencies

MindsDB integrates with numerous data sources and AI frameworks. To use any of the integrations, you should ensure that the required dependencies are installed in the Docker container.

Method 1

Install dependencies directly from MindsDB editor. Go to Settings and Manage Integrations, select integrations you want to use and click on Install.

Method 2

Start the MindsDB Docker container:

bash

docker start mindsdb_container

<Tip> If you haven't specified a container name when spinning up a container with `docker run`, you can find it by running `docker ps`. </Tip> <Tip> If you haven't yet created a container, use this command:

bash

docker run --name mindsdb_container -e MINDSDB_APIS=http -d -p 47334:47334 mindsdb/mindsdb

</Tip>

Start an interactive shell in the container:

bash

docker exec -it mindsdb_container sh

Install the dependencies:

bash

pip install .[handler_name]

For example, run this command to install dependencies for the OpenAI handler:

bash

pip install .[openai]

Exit the interactive shell:

bash

exit

Restart the container:

bash

docker restart mindsdb_container

Configuration

This is a configuration for MindsDB's Docker image that includes storage location, log level, debugging information, installed integrations, and API endpoints. These parameters can be customized by modifying a JSON file that stores default configuration.

Default configuration

The default configuration for MindsDB's Docker image is stored as a JSON code, as below.

json

{
    "config_version":"1.4",
    "paths": {
        "root": "/root/mdb_storage"
    },
    "debug": false,
    "integrations": {},
    "api": {
        "http": {
            "host": "0.0.0.0",
            "port": "47334"
        },
        "mysql": {
            "host": "0.0.0.0",
            "password": "",
            "port": "47335",
            "user": "mindsdb",
            "database": "mindsdb",
            "ssl": true
        }
    }
}

Custom configuration

To override the default configuration, you can mount a config file created in your host machine over /root/mindsdb_config.json, as below.

bash

docker run --name mindsdb_container -e MINDSDB_APIS=http -d -p 47334:47334 -v $(pwd)/mdb_config.json:/root/mindsdb_config.json mindsdb/mindsdb

<Tip> **What's next?**

Now that you installed and started MindsDB locally in your Docker container, go ahead and find out how to create and train a model using the CREATE MODEL statement.

Check out the Use Cases section to follow tutorials that cover Large Language Models, Chatbots, Time Series, Classification, and Regression models, Semantic Search, and more. </Tip>