Delete one S3 object safely

Part 20 of AWS from Zero. Today we delete exactly one object.

What we are learning

For a non-versioned bucket, deleting an object is permanent. A careful command-line flow is: inspect, back up when needed, delete, then verify.

Before you run it

REGION="ap-south-1"
BUCKET="aws-zero-delete-one-12345"
KEY="notes/important.txt"
 
aws s3api create-bucket \
  --bucket "$BUCKET" \
  --region "$REGION" \
  --create-bucket-configuration LocationConstraint="$REGION"
 
echo "important demo file" > important.txt
aws s3 cp important.txt "s3://$BUCKET/$KEY"

Inspect the exact key first:

aws s3api head-object --bucket "$BUCKET" --key "$KEY"

Optional backup

aws s3 cp "s3://$BUCKET/$KEY" important-backup.txt

Delete the object

aws s3api delete-object \
  --bucket "$BUCKET" \
  --key "$KEY"

Verify the result

aws s3api list-objects-v2 \
  --bucket "$BUCKET" \
  --prefix "$KEY" \
  --query "Contents[].Key" \
  --output text

No returned key means the current object is no longer listed.

One tiny variation

The high-level equivalent is shorter:

aws s3 rm "s3://$BUCKET/$KEY"

Use s3api delete-object when you need low-level options such as a specific version ID.

Common mistake

S3 delete operations are idempotent for a missing key: deleting a key that is already absent can still return success. Always verify the exact bucket and key before and after deletion.

Versioned buckets behave differently. A normal delete usually creates a delete marker instead of permanently removing old data. We will see that shortly.

Cleanup

aws s3api delete-bucket --bucket "$BUCKET" --region "$REGION"
rm important.txt important-backup.txt

Next, we will delete several known keys in one request.

AWS CLI delete-object reference