Your Drupal deploy should never take the site down. But most teams still deploy like this: push code, run drush updb, import config, clear cache, and pray.
I did that too - until a bad deploy took down a regulated, always-on platform at 2am.
In this session, I'll show you the exact blue-green pipeline I now use for Drupal: build once, deploy to green, run database updates + config import safely, run automated smoke tests on real pages, flip traffic, and keep blue ready for instant rollback. You'll see:
- How to handle drush deploy, config sync, and database updates without locking editors out
- How to invalidate cache and warm Varnish/CDN without a stampede
- How to run content-freeze vs. no-freeze deploys - and when to use each
- A live demo of a failed deploy rolling back in under 60 seconds
You'll leave with a 1-page deploy checklist and a sample GitHub Actions / GitLab CI pipeline you can copy for Pantheon, Acquia, AWS, or self-hosted Drupal. This is for Drupal developers, DevOps engineers, and tech leads who are tired of risky Friday deploys. What you will learn:
- Design a zero-downtime deploy for Drupal 10/11
- Run config + database changes safely while editors are working
- Build and test a rollback before you actually need it