Creating a Website on S3

Task 1: Connecting to an Amazon Linux EC2 Instance

First, I needed to establish a connection to my EC2 instance. I clicked the Details button at the top of the lab interface and selected Show to reveal the InstanceSessionUrl. After copying this URL, I pasted it into a new browser tab which opened a console connection to the instance using AWS Systems Manager Session Manager.

Once connected, I ran these commands to switch to the ec2-user and verify my home directory:

Task 2: Configuring the AWS CLI

The AWS CLI was already pre-installed on the Amazon Linux instance, which saved me some time. I ran the configuration command to set up my credentials:

At the prompts, I entered:

• AWS Access Key ID: I copied and pasted the AccessKey value from the lab environment
• AWS Secret Access Key: I copied and pasted the SecretKey value from the lab environment
• Default region name: us-west-2
• Default output format: json

Task 3: Creating an S3 Bucket

Next, I needed to create an S3 bucket with a unique name. I decided to follow the naming convention of combining my first initial, last name, and three random numbers. I executed this command to create my bucket:

After running the command, I received a JSON response confirming successful creation with the location URL: http://askeries754.s3.amazonaws.com/

Task 4: Creating an IAM User with S3 Access

I needed to create a new IAM user with full access to Amazon S3. I did this by running:

Then I created a login profile for this user:

To sign in as this user in the AWS Console, I needed the account number. I found this by:

1. Clicking on the User dropdown menu in the top right corner of the AWS Management Console
2. Copying the 12-digit Account ID
3. Signing out of the current session which I figured out caused extra unnecessary steps of signing back in again so I opened a private window to sign in instead but also using multi-session option in User dropdown could be a valid alternative option in this case
4. Signing in as an IAM user using the account ID
5. Entering awsS3user as the username and Training123! as the password

Once signed in as awsS3user, I navigated to the S3 console and discovered I couldn't see my bucket yet because I hadn't granted the necessary permissions.

Back in my terminal, I ran this command to find the appropriate policy for S3 full access:

After finding the policy name in the results (AmazonS3FullAccess), I attached it to my user:

I then refreshed the browser tab in the AWS Console and could now see my S3 bucket.

Task 5: Adjusting S3 Bucket Permissions

To make my bucket publicly accessible for website hosting, I needed to adjust the permissions:

1. In the S3 console, I selected my bucket and went to the Permissions tab
2. Under "Block public access", I clicked Edit
3. I deselected "Block all public access"
4. I confirmed the changes when prompted
5. Under "Object Ownership", I clicked Edit
6. I selected "ACLs enabled"
7. I acknowledged that ACLs would be restored
8. I saved the changes

Task 6: Extracting the Website Files

Now I needed to extract the static website files that would be uploaded to S3. In the SSH terminal, I ran:

To verify the extraction was successful, I ran ls and confirmed I could see the index.html file and the css and images directories.

Task 7: Uploading Files to Amazon S3

With my files ready, I needed to configure my bucket for website hosting and upload the files:

First, I configured the bucket as a website and set the index document:

Then I uploaded all the files with public read access:

To verify the upload was successful, I listed the contents of my bucket:

In the AWS Console, I went to my bucket's Properties tab and noticed that Static website hosting was enabled. I clicked on the Bucket website endpoint URL and saw my Café & Bakery website successfully deployed and accessible to the public!

Task 8: Creating a Batch File for Website Updates

To make future website updates easier, I created a batch file:

1. I checked my command history to find the S3 copy command I used earlier:
history
2. I created a new shell script:
cd ~
touch update-website.sh
3. I opened the file in the VI editor:
vi update-website.sh
4. I added the following content (pressing 'i' to enter insert mode):
#!/bin/bash aws s3 cp /home/ec2-user/sysops-activity-files/static-website/ s3://askeries754/ --recursive --acl public-read
5. I saved the file (pressing Esc, then typing :wq and Enter)
6. I made the script executable:
chmod +x update-website.sh

To test the script, I made some changes to the website:

1. I opened the index.html file:
vi sysops-activity-files/static-website/index.html
2. I changed the background colors:
• Changed the first instance of bgcolor="aquamarine" to bgcolor="gainsboro"
• Changed bgcolor="orange" to bgcolor="cornsilk"
• Changed the second instance of bgcolor="aquamarine" to bgcolor="gainsboro"
3. I saved the file (Esc, :wq, Enter)
4. I ran my update script:
./update-website.sh

I refreshed the website in my browser and saw the color changes, confirming my update process worked!

Optional Challenge: Making the Update Process More Efficient

I noticed that my batch file was uploading all files every time, even when only one file changed. To optimize this, I explored the AWS CLI's sync command.

1. I made another small change to index.html
2. Instead of using aws s3 cp, I updated my script to use aws s3 sync:
aws s3 sync /home/ec2-user/sysops-activity-files/static-website/ s3://askeries754/ --acl public-read

When I ran the updated script, I noticed it only uploaded the changed file (index.html) rather than all files. This made the update process much more efficient by:

• Only transferring files that had changed
• Reducing data transfer and bandwidth usage
• Making updates faster, especially for larger websites
• Maintaining better synchronization between local and remote versions

The sync command was clearly more efficient than the cp command for managing website updates.