---
title: Connect a Data Source
sidebarTitle: Connect a Data Source
---

## Description

MindsDB lets you connect to your favorite databases, data warehouses, data lakes, etc., via the `CREATE DATABASE` command.

The MindsDB SQL API supports creating connections to integrations by passing the
connection parameters specific per integration. You can find more in the
[Supported Integrations](#supported-integrations) chapter.

<Note>
MindsDB doesn't store or copy your data. Instead, it fetches data directly from your connected sources each time you make a query, ensuring that any changes to the data are instantly reflected. This means your data remains in its original location, and MindsDB always works with the most up-to-date information.
</Note>

## Syntax

Let's review the syntax for the `CREATE DATABASE` command.

```sql
CREATE DATABASE [IF NOT EXISTS] datasource_name
[WITH] [ENGINE [=] engine_name] [,]
[PARAMETERS [=] {
  "key": "value",
  ...
}];
```

On execution, we get:

```sql
Query OK, 0 rows affected (x.xxx sec)
```

Where:

| Name                | Description                                                                       |
| ------------------- | --------------------------------------------------------------------------------- |
| `datasource_name`   | Identifier for the data source to be created.                                     |
| `engine_name`       | Engine to be selected depending on the database connection.                       |
| `PARAMETERS`        | `{"key": "value"}` object with the connection parameters specific for each engine. |

<Note>
**SQL Commands Resulting in the Same Output** Please note that the
keywords/statements enclosed within square brackets are optional. Also, by
default, the engine is `mindsdb` if not provided otherwise. That yields the
following SQL commands to result in the same output.

```sql
CREATE DATABASE db;
CREATE DATABASE db ENGINE 'mindsdb';
CREATE DATABASE db ENGINE = 'mindsdb';
CREATE DATABASE db WITH ENGINE 'mindsdb';
CREATE DATABASE db USING ENGINE = 'mindsdb';
```

</Note>


### What's available on your installation

<Info>
Here is how you can query for all the available data handlers used to create database connections.

  ```sql
  SELECT *
  FROM information_schema.handlers
  WHERE type = 'data';
  ```

  Or, alternatively:

  ```sql
  SHOW HANDLERS
  WHERE type = 'data';
  ```

And here is how you can query for all the connected databases:

  ```sql
  SELECT *
  FROM information_schema.databases;
  ```

  Or, alternatively:

  ```sql
  SHOW DATABASES;

  SHOW FULL DATABASES;
  ```
</Info>


## Example

### Connecting a Data Source

Here is an example of how to connect to a MySQL database.

```sql
CREATE DATABASE mysql_datasource
WITH ENGINE = 'mariadb',
PARAMETERS = {
  "user": "root",
  "port": 3307,
  "password": "password",
  "host": "127.0.0.1",
  "database": "my_database"
};
```

On execution, we get:

```sql
Query OK, 0 rows affected (8.878 sec)
```

### Listing Linked Databases

You can list all the linked databases using the command below.

```sql
SHOW DATABASES;
```

On execution, we get:

```sql
+--------------------+
| Database           |
+--------------------+
| information_schema |
| mindsdb            |
| files              |
| mysql_datasource   |
+--------------------+
```

## Making your Local Database Available to MindsDB

When connecting your local database to MindsDB Cloud, you should expose the
local database server to be publicly accessible. It is easy to accomplish using
[Ngrok Tunnel](https://ngrok.com). The free tier offers all you need to get
started.

The installation instructions are easy to follow. Head over to the
[downloads page](https://ngrok.com/download) and choose your operating system.
Follow the instructions for installation.

Then [create a free account at Ngrok](https://dashboard.ngrok.com/signup) to get
an auth token that you can use to configure your Ngrok instance.

Once installed and configured, run the following command to obtain the host and
port for your localhost at `port-number`.

```bash
ngrok tcp port-number
```

Here is an example. Assuming that you run a PostgreSQL database at
`localhost:5432`, use the following command:

```bash
ngrok tcp 5432
```

On execution, we get:

```bash
Session Status                online
Account                       myaccount (Plan: Free)
Version                       2.3.40
Region                        United States (us)
Web Interface                 http://127.0.0.1:4040
Forwarding                    tcp://4.tcp.ngrok.io:15093 -> localhost 5432
```

Now you can access your local database at `4.tcp.ngrok.io:15093` instead of
`localhost:5432`.

So to connect your local database to the MindsDB GUI, use the `Forwarding`
information. The host is `4.tcp.ngrok.io`, and the port is `15093`.

Proceed to create a database connection in the MindsDB GUI by executing the
`CREATE DATABASE` statement with the host and port number obtained from
Ngrok.

```sql
CREATE DATABASE psql_datasource
WITH ENGINE = 'postgres',
PARAMETERS = {
  "user": "postgres",
  "port": 15093,
  "password": "password",
  "host": "4.tcp.ngrok.io",
  "database": "postgres"
};
```

Please note that the Ngrok tunnel loses connection when stopped or canceled. To
reconnect your local database to MindsDB, you should create an Ngrok tunnel
again. In the free tier, Ngrok changes the host and port values each time you
launch the program, so you need to reconnect your database in the MindsDB Cloud
by passing the new host and port values obtained from Ngrok.

Before resetting the database connection, drop the previously connected data
source using the `DROP DATABASE` statement.

```sql
DROP DATABASE psql_datasource;
```

After dropping the data source and reconnecting your local database, you can use
the predictors that you trained using the previously connected data source.
However, if you have to `RETRAIN` your predictors, please ensure the database
connection has the same name you used when creating the predictor to avoid
failing to retrain.

## Supported Integrations

The list of databases supported by MindsDB keeps growing. Check out all our [database integrations here](/data-integrations/all-data-integrations).