Deploy Schema Change with Rollout Policy

​

Features covered

• Level 1 – Automatic rollout (Rollout Policy) (available in all plans)
• Level 2 – Manual rollout with SQL Review (available in all plans)
• Level 3 – Custom Approval (available in Enterprise)

​

Preparation

1. Ensure Docker is running, then start the Bytebase container:
   Copy

   Ask AI

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

   
2. Open Bytebase in localhost:8080, fill in the fields and click Create admin account. You’ll be redirected to Workspace.
3. During workspace setup, choose the built-in sample data.

​

Level 1: Automatic rollout in the Test environment (All plans)

1. In Workspace, go to Environments > Test, and check Automatic under the Rollout policy.
2. From the top-left project switcher, select Sample Project. Under the CI/CD section, click Plans, then + New Plan. Choose Schema Migration (DDL) and select the hr_test database.
3. Using the Schema Editor, add a new column nickname to the employee table and click Confirm.
4. Click Create, then click Ready for Review.
5. The rollout executes automatically to the Test environment. The issue is marked as Done.

​

Level 2: Manual rollout with SQL Review (All plans)

1. In Workspace, go to Environments > Prod. Leave the Automatic unchecked under the Rollout policy. Keep the default setting without changing, it’s Require Issue Approval and Plan Check Enforcement>Block on Errors Only.
2. We’ll add an SQL Review policy to the Prod environment. Click SQL Review Policy to open SQL Review under the Environment > Prod. Select the Sample Template, which has 20 enabled rules.
3. Click Confirm to add the SQL Review policy to the Prod environment.
4. From the top-left project switcher, select Sample Project. Under the CI/CD section, click Plans, then + New Plan. Choose Schema Migration (DDL) and select the hr_prod database.
5. Using the Schema Editor, add a new column nickname to the employee table and click Confirm.
6. Intentionally violate the NOT NULL rule. Click Create; SQL Review runs automatically and shows a warning for the Prod task.
7. Click the warning to view details.
8. Go to the Rollout tab, click Run. You will see a warning that the SQL Review is not passed. However, you can bypass the SQL Review by checking Bypass to continue rollout.
9. Go back to the Prod environment, and set Plan Check Enforcement > Strict Enforcement under the Rollout Policy, then come back to the plan and refresh, you will see this time you can’t bypass the SQL Review.
10. Fix the SQL by adding the NOT NULL constraint. Click Save to update the plan. SQL Review runs again and should pass. Click Ready for Review to submit the plan for approval. Because here we don’t have approval flow configured, the plan will be approved automatically.
11. The rollout is now ready, click Run again, this time you will see the SQL Review is passed.
12. After the rollout completes, the issue is marked as Done.

​

Level 3: Manual rollout with custom approval (Enterprise)

​

Summary

• GitOps: Observe Git pushes and trigger schema migrations.
• Batch changes: Change multiple databases in a single workflow.