DbSchema Databricks ER Diagram Designer

DbSchema is a Databricks ER diagram database designer and query tool. Using DbSchema, you can design and document Databricks databases while connected to the database server or without a connection.

Next, you can share the design with the team using Git. The design model loaded in DbSchema can be compared with the database, synchronization scripts can be generated, and the schema can be deployed to any database server.

Download DbSchema

How to Connect to Databricks Database

1. Access the Connection Dialog

   Choose "Connect to the database" or "New Model Connected to the Database" will let you select your database and open the Connection Dialog.

   DbSchema automatically downloads the JDBC driver to connect to your database.

2. Configure the Connection

In the Connection Dialog window, select the JDBC URL you want to connect with. If your database is running in the cloud, choose Edit Manually and insert the JDBC URL from the cloud console.

Enter the host name, the authentication details, select your database as described in the Connection Dialog page.

3. Troubleshoot Firewall Issues
If you encounter connection issues, it may be due to firewall settings. Read how to enable Firewall Connections.

Getting the JDBC URL

First, you have to create a Databricks in Azure and then use 'Launch Workspace' to access it. If you have multiple Compute items, click one of them.

In the cluster configuration scroll on the bottom to the 'Advanced' section.

There click the Advanced Options. There you will find the JDBC URL:

In DbSchema choose to manually connect to the database, and copy the JDBC URL. In the URL you have to replace the TOKEN as described below.

When connecting, you can leave the user and password fields empty.

You can test this using the command `openssl s_client -connect databricks-hostname:443`. On Windows, this needs to be run from Git Bash ( from Windows Git client ), not from command prompt.

Required Privileges

You should have the 'USE' privileges on the catalog you want to reverse engineer. The 'BROWSE' privileges are not enough.

Getting the Access Token

The token generation is documented as described in the Databricks documentation.

• In your Azure Databricks workspace, click your Azure Databricks username in the top bar, and then select User Settings from the drop down.

  

• On the Access tokens tab, click Generate new token.

Generate an access token:

Making sure the compute nodes have enough quota

To be able to start a compute node, you need to make sure that the particular compute node type you want to use has enough quota.

For example, let's say you want to run the compute resource on a node from the family Standard DS (version 2). Note that you need at least 4 cores (vCPU-s):

To make sure this node type family has enough vCPU quota, do the following:

1. go to the Azure Portal home
2. go to subscriptions
3. click on the subscription that is linked to the compute resource you want to start
4. go to Settings and click on "Usage and quotas"
5. find your family in the list and click on it
6. click on "New Quota Request"
7. choose the new quota (in this case we need at least 4 vCPU-s), and then click "Submit":
   
8. Wait for the request to be successful: