🚀 How to Start Any Infrastructure Upgrade (A Simple Checklist)

🚀 How to Start Any Infrastructure Upgrade (A Simple Checklist)

Upgrades don’t fail because of tools. They fail because people start without understanding what already exists.

Here’s a simple, proven way to begin any infrastructure upgrade — especially if you’re early in your career.

1. Start With the Current State

Before changing anything, list what’s already deployed:

terragrunt/ terraform state list

This tells you: What resources exist, What Terraform manages, What must still exist after the upgrade

👉 This is your baseline.

2. Understand the System Roles

Always verify:

• Which node is leader
• Which nodes are followers
• How DNS (Route53) maps to instances

For distributed systems, upgrading in the wrong order can cause outages.

3. Confirm DNS & Ownership

Check Route53 and decide:

• Import existing records if they already exist
• Create only if they don’t

Never assume Terraform ownership — always verify.

4. Save Evidence Before & During Upgrade

For each node:

• Save Terraform plan output
• Capture resources before and after
• Note any warnings or expected failures

Even a simple screenshot or note is enough.

5. Upgrade One Thing at a Time

Upgrade:

• One node
• Validate it fully
• Then move to the next

1) Exhibitor UI (health/status)

http://zookeeper-kafka-00x-<env>.pvt:8090

2) DNS resolves to the expected IP

dig +short zookeeper-kafka-00x-<env>.pvt

3) Exhibitor logs

journalctl -u exhibitor -f

4) ZooKeeper role (leader/follower)

echo srvr | nc zookeeper-kafka-00x-<env>.pvt 2181

This keeps rollback safe and risk low.

6. Document Everything (Even “No Issues”)

Write down:

• What worked
• What was expected
• Any issues — or clearly state no issues

This helps future engineers repeat the process confidently.

Final Thought

A good upgrade is not just successful — it’s repeatable.

If someone new can follow your notes and do it again, you’ve done the job right. All the best!…