Manual AMI Upgrades? That’s History Now.

We used to spend undifferentiated hours manually upgrading AMIs:
• Jumping on calls
• Identifying which MongoDB node was a secondary
• Carefully upgrading one by one, hoping not to cause downtime

Now?

We just head to GitHub Actions, select the stack name, provide the new AMI ID, and hit Run.
That’s it.

The pipeline does the rest:
• Discovers all nodes in the stack
• Identifies secondaries in our multi-node MongoDB cluster
• Upgrades one secondary at a time
• Runs post-upgrade health checks
• Waits for manual approval before touching the primary

It’s cautious. It’s automated. And it’s reliable.

We’ve codified our operational knowledge into a pipeline—what used to be tribal knowledge and Slack threads is now reproducible, trackable, and safe.

The impact?
• ✅ No downtime
• ✅ Reduced operational risk
• ✅ Engineers spend time solving real problems—not babysitting upgrades
• ✅ Clear audit trail & GitOps-friendly

This isn’t just Infra as Code—
This is Operations as Code.

What part of your infra still feels like a manual chore?