Database CI/CD and Schema Migration with PostgreSQL and GitHub

ngrok is a reverse proxy tunnel, and in our case, we need it for a public network address in order to receive webhooks from VCS. ngrok we used here is for demonstration purposes. For production use, we recommend using Caddy.

1. Run Bytebase in Docker with the following command:

   docker run --rm --init \
     --name bytebase \
     --publish 8080:8080 --pull always \
     --volume ~/.bytebase/data:/var/opt/bytebase \
     bytebase/bytebase:3.1.0

2. Bytebase is running successfully in Docker, and you can visit it via localhost:8080. Register an admin account and it will be granted the workspace admin role automatically.

3. Login to ngrok Dashboard and complete the Getting Started steps to install and configure. If you want to use the same domain each time you launch ngrok, go to Cloud Edge > Domains, where you'll find the domain <<YOURS>>.ngrok-free.app linked to your account.

4. Run the ngrok command ngrok http --domain=<<YOURS>>.ngrok-free.app 8080 to start ngrok with your specific domain, and you will see the output displayed below:

   

5. Log in Bytebase and click the gear icon (Settings) on the top right. Click General under Workspace. Paste <<YOURS>>.ngrok-free.app as External URL under Network section and click Update.

   

6. Now you can access Bytebase via <<YOURS>>.ngrok-free.app.

1. Go to Bytebase homepage, and click Integration > GitOps on the left sidebar. Choose GitHub.com as Git provider. What we need is a github personal access token.

2. Go to your GitHub account. Click your avatar and then click Settings on the menu. Click Developer settings on the left sidebar, and then click Personal access tokens > Fine-grained tokens.

3. Click Generate new token, fill in the fields and check the scopes according to the description on Bytebase. Click Generate token.

4. Copy the token and paste it back into Bytebase Integration > GitOps. Click Confirm and add.

1. Go to the project Sample Project, click Integration > GitOps. Click Add Enable GitOps connector.

2. Choose GitHub.com - the provider you just added. It will display all the repositories you can manipulate. Choose test-bb-gitops.

3. Keep the default setting, and click Finish. Pay attention to Database Group, which is the database group that the schema change will be applied to. With Community Plan, the changes will automatically affect all databases in the project. With Enterprise Plan, you'll have the option to specify the target database group.

1. In your GitHub repository test-bb-gitops, create a folder bytebase, then create an sql file 202404151600_create_table_t1.sql.

   Paste the sql script in it.

   CREATE TABLE t1 (
      "id" INTEGER NOT NULL
   );

2. Create a new branch for this commit and start a pull request. Click Merge pull request to merge the new branch into the main branch.

3. There will be a comment saying there's a rollout in Bytebase. Click and go to the issue page, you’ll see

   1. The issue is created via GitHub.com, there's a link to the GitHub commit.

   2. The SQL is exactly the one we have committed to the GitHub repository.

   3. The SQL has passed the automatic task checks and rollout automatically.

   4. Since there're two databases in the project, Bytebase creates a 2-staged pipeline to roll out the change sequentially.

      

4. Click View change, you can view the schema diff.

Now you have tried out GitOps workflow, which will store your database schema in GitHub and trigger the change upon committing the change to the repository via Pull Request, to bring your database change workflow to the next level of Database DevOps - Database as Code.

If the built-in workflow is not suitable, you can opt to Bytebase API to fully customize the workflow to integrate with your CI pipeline. Automating Database Schema Change workflow Using GitHub Actions is an example.