S3 Meets My Terminal – CLI & Versioning Experiments

This ‘day 14’ is almost a week later, reason being I was stuck with some critical projects assigned by my great manager which took me away from practicing the AWS. Got time today and did hands dirty on AWS S3.

After a full day of clicking around the S3 console, I figured it was time to stop being a tourist and actually get comfortable with the AWS CLI.

I mean, real engineers use the terminal, right? 😅
So today was all about interacting with S3 using the CLI and then diving into one of its coolest features: versioning.

---

🔧 Setting Up the CLI (If You Haven’t Already)

First, I checked if I had the AWS CLI installed:

aws --version

If you don't have it yet, install it using:

# On Ubuntu/Debian
sudo apt install awscli

# On macOS (with Homebrew)
brew install awscli

Then I configured it with my IAM credentials:

aws configure

It asked for:

• AWS Access Key ID

• AWS Secret Access Key

• Region (I used us-east-1)

• Output format (I chose json)

Boom. CLI ready.

🪣 Creating a Bucket from Terminal

I wanted to do everything via CLI today, so I started by creating a new S3 bucket:

aws s3 mb s3://cli-demo-bucket-username

✅ Success.

Then I listed all my buckets:

aws s3 ls

Seeing your bucket show up in the list? That’s when you know it worked.

Uploading Files via CLI

I created a simple text file:

echo "First version of this file" > notes.txt

And uploaded it:

aws s3 cp notes.txt s3://cli-demo-bucket-username

Then updated the file:

echo "Second version – updated content" > notes.txt
aws s3 cp notes.txt s3://cli-demo-bucket-username

At this point, I had overwritten the file in S3 — and that's where versioning comes into play.

Enabling Versioning on a Bucket

I wanted to keep track of every version of notes.txt. So I turned on versioning:

aws s3api put-bucket-versioning --bucket cli-demo-bucket-username \
--versioning-configuration Status=Enabled

To verify:

aws s3api get-bucket-versioning --bucket cli-demo-bucket-username

Now, every time I upload a file with the same name, S3 keeps the older versions. No more overwriting and losing changes.

Seeing All Versions of an Object

After modifying notes.txt a couple of times, I ran:

aws s3api list-object-versions --bucket cli-demo-bucket-username

And yup — I could see multiple VersionIds for notes.txt.
Each upload has its own unique ID, and I can retrieve or delete a specific version whenever I want.

Fun Experiment: Restore an Old Version

Just to try it out, I downloaded an older version like this:

aws s3api get-object --bucket cli-demo-bucket-username \
--key notes.txt \
--version-id <OLD_VERSION_ID> \
restored.txt

It worked like a charm. My terminal became a time machine for files.

---

🧠 What I Learned Today

• Using S3 from the CLI feels powerful — way faster than clicking through the console

• Versioning is a lifesaver — especially when you’re dealing with frequent updates

• AWS CLI responses are super detailed — helpful for debugging and scripting later

• Every new feature I touch in AWS seems simple at first, but always has depth if you look closer

Up Next

I’m thinking of continuing the S3 journey with:

• Lifecycle rules (automatically moving or deleting old versions)

• Trying public vs private uploads via CLI

• Maybe even setting up a basic backup system using S3 and versioning

This storage service is turning out to be way more than a file bucket.
Hopefully, I am not caught with some “critical company work…..”