Pipeline Operation: Rollback

This page explains how to configure, run, and manage a Rollback operation in Revvy pipelines. Rollback lets you revert changes applied by a previous deployment, restoring the org to its earlier state.

The Rollback operation is linked to a specific Deploy step. When a deployment encounters issues (e.g., failed tests, broken dependencies), you can automatically or manually roll it back.

Rollback ensures stability by:

  • Reverting changes made during a deployment.

  • Protecting higher environments from breaking changes.

  • Allowing automated “fail-safe” paths in pipelines.

Creating a Rollback Step

  1. Open or create a pipeline with an existing Deploy step.

  2. Click the + button after the Deploy node.

  3. In the configuration panel:

    • Name: Enter a step name (e.g., Rollback after Deploy).

    • Operation: Select Rollback.

    • Deployment: Choose the Deploy step to link this rollback to.

    • Next step:

      • On pass → Stop or proceed to another step.

      • On fail → Stop or retry strategy.

  4. Click Add.

Example: A Rollback step was added after Deploy to Sandbox, configured to Stop regardless of outcome.

Running a Rollback

  1. Run the pipeline normally.

  2. If the Deploy step fails or is configured to trigger rollback, the pipeline automatically executes rollback.

  3. The pipeline diagram will show a red line for failure and a green line when rollback completes successfully.

Rollbacks can also be run manually by executing the pipeline section containing the Deploy + Rollback sequence.

Editing a Rollback Step

  • Select the Rollback node in your pipeline diagram.

  • Click Edit to update:

    • Linked deployment step.

    • Pass/fail next steps.

  • Save changes to apply updates.

This allows you to change rollback behavior, for example, making it conditional only on certain types of failures.

Viewing Rollback Results

After execution, you can view rollback status under Activities:

  • Jobs:

    • Shows Rollback type with status, start/end time, and duration.

    • Example: Rollback completed in 10 seconds after a failed deploy.

  • Pipelines:

    • Displays the full flow including Deploy → Rollback → Stop.

  • Logs:

    • Detailed system messages such as: [info] Prepare destructive changes started

      [info] Salesforce metadata deployment started

      [info] Rollback completed

Best Practices

  • Always link rollback to a specific deployment for traceability.

  • Use rollback as the On fail step for production deployments.

  • Review logs carefully to ensure the rollback reverted all intended changes.

  • Combine rollback with notification steps (future feature) to alert teams when it is triggered.

Last updated

Was this helpful?