Why Amazon Aws S3 is Better for The Static Web Hosting

Amazon is the world’s largest online retail giant. Jeff Bezos incorporated the company on July 1994, by launching website to the public a year later. The original name was to be “Cadabra”, but Jeff Bezos mistakenly heard it as “Cadaver.” The company before starting into online retail business was mostly focused on selling books first which was being sold out of Bezos’ garage. Today the company is internationally carrying out the business. It sells consumers electronic devices, the company provides compute, storage database and other AWS Services as well as fulfillment and publishing the digital content subscriptions.

Amazon Web Services (AWS) was launched in 2006 year from the internal infrastructure that amazon.com established to handle their online retail operations. It was the first company who offered pay-as-you-go cloud computing models which provides users with compute, storage services. AWS offers services form dozens of the data centers which are available across the global which are called availability zones (AZs) across the world. AZs is represented as multiple physical data centers and a region is a called collection of AZs in geographical proximity which are connected by the low-latency network links. The customers of AZs can have the virtual machines replicate the data in the different zones of AZs to provide better infrastructure which can be resistant to the failure of the particular servers or the entire data center. It provides more than 100 services which includes as follows:

• Compute
• Databases, data management
• Storage
• Networking
• Migration, hybrid cloud
• Management, monitoring
• Development tools and application services
• Big data management, analytics
• Security, governance
• Artificial Intelligence
• Mobile Development
• Messages, notifications
• Other services like Amazon WorkDocs, Amazon WorkSpaces, Amazon AppStream, Amazon IoT Button, AWS Greengrass.

The primary focus of this project was about the storage and compute service. Amazon Simple Storage Service(S3) provides the scalable object storage for the data backups and analytics. Business can cut cost of the money with the use of the S3 for example an IT Administrator can stores it data and files as S3 objects which can be inside the S3 buckets to keep them in well-organized manner. Amazon Elastic Compute Cloud (EC2) helps to provide virtual servers which are called instances for the computing capacity. In this report I will collate between two of the above services EC2 and S3.

Materials

In this portion of the project I will elaborate the materials being used for the project which made possible to achieve the goal of the project and provide their depiction as follows:

• Amazon Web Service Management Console. It a browser-based GUI for the AWS to handle the services.
• Terminal (MAC OS). It is similar like CMD which is being used in Windows OS which helps to connect through Secure Shell (SSH) into the Centos OS.
• Cisco AnyConnect Secure Mobility Client. It is a program which helps to connect to campus server, so we can access the virtual machine data.
• Centos OS virtual machine which is hosted on the coivcenter1.hh.nku.edu server. It is Linux based Operating system which is being hosted on the coivecenter1.hh.nku.edu.
• Amazon Web Service Command Line Interface (CLI). AWS services are being managed and configured by the CLI which is called command line tool.
• Apache HTTP Sever on Amazon Elastic Cloud (EC2) ubuntu instance. It is web server which helps to serve HTTPS and HTTP requests.
• Amazon Web Service Elastic Cloud (EC2) and EC2 ubuntu instance. It is Linux a Virtual Machine hosted on the Amazon Web Service (AWS).

Amazon Simple Service Storage S3 buckets. It is storage which is being hosted on Amazon Web Service. Amazon Monthly Calculator. It is a calculator used to measure the cost associated with the services. Google Chrome Web Browser. Used to run the test hosted on the web. Microsoft Excel. It is being used to analyze the data being collected and provide graphical visual information.  Apache HTTP Benchmark tool, ab.

Method

In this part of the project I will review about the completion of this project by using the materials as described in above part of the project. To achieve the final outcome of the project I have carefully went through steps as follows: -

Navigate to your Web Browser and visit https://console.aws.amazon.com/ webpage and create free tier account and login with credentials once finished with signing up for the account.

When you are successfully logged into the console, navigate to the “AWS Services” search tab and search out for the “IAM” service and click on it as seen in below picture.

You will see the following screen when you navigate to IAM. It is an Identity and Access Management Page. It will help you to maintain and configure the user access to your account.

As you see in the above picture on your left side as you see the navigation bar as arrowed click on the “Users”. There will be a new webpage which will display two options as “Add User” and “Delete User”. Click on the “Add User” button and it will navigate you towards adding new user to the webpage as seen in below picture. Once you fill up the HTML form which is being displayed on the web page please remember to check the boxes named Programmatic access and AWS Management Console Access and Custom Password. Uncheck the box which says Request password reset and then navigate to the next: permission button.

As you navigate to the next page it will ask you to select the permissions for the user. Please be advised if there is no user group created then you will need to create group for the user first. As you see in the below image there is group created so now just select the checkbox under the select policu to set the permission boundary with Administartor Access and select the group which was created and naviagate to next:review button.

When you are navigated to the next page it will ask you review the details provided before. Finally, when you have reviewed then click “Create user”. It will show Success message box and there will be button to download the .csv file. The .csv file will contain important Access key ID and your Secret Access key which plays important role in linking your account to AWS account with help of the AWS CLI tool.

Open the terminal window [ If you are using MAC OS else from Windows OS open CMD]. Connect to NKU Centos OS virtual machine through SSH connection. Please be advised to be connected through NKU’S Network otherwise you won’t be able to access it.

To access the AWS CLI tool, you will be needing the root access so change to root access. Enter the “aws configure” command to configure the AWS CLI tool so that you can connect to your AWS account. It will ask you to enter the AWS Access Key ID, then to enter the AWS Secret Access Key, your region name and output. Those data will be in your .csv file.

Next step is to create the EC2 instance. You will need to create security group to allocate to EC2 before creation of the EC2 instance. It can be done by entering command as seen in the image posted below:

You will need to add rules to the security group by entering following commands as seen in the below image:

[You will need two different rules for two different ports one for SSH (22) and another one for the HTTP (80).

To verify if the rules are been added to your security group you can use the command “aws ec2 describe-security-groups --group-names your_username”.

When you are connecting to the EC2 instance in order to get connected you will need key pair and to do that you will use command “aws ec2 create-key-pair --key-name your_username-key --query 'KeyMaterial' --output text > devenv-key.pem”.

In order to make your file accessible only to you will need to modify the permission of the file by entering command “chmod 400 devenv-key.pem(Your Key .pem)”. Please be advised that by not changing the permission it won’t let enter into the EC2 instance.

To make the instance run you enter command aws ec2 run-instances --image-id ami-d9a98cb0 --count 1 --instance-type t1.micro --key-name your_username-key --security-groups your_username”. [ By not taking down the details of the instance ID it will result in loss of the achieving the final goal of the project].

You will need the public IP address or the public DNS name of the instances to get the access of the instance. By waiting for about 30 seconds you run the command “aws ec2 describe-instances --instance-ids your_instance_id” ( To get the network related information). As seen in the below image.

To connect to the instance, you will ssh command ssh -i devenv-key.pem ubuntu@the_public_IP_address_of_your_instance. It will prompt up message by asking Are you sure you want to continue connecting (yes/no)?”, please type “Yes” to continue. When you are been promoted then type “sudo -s” to access in the root mode.

To run the apache web server on your EC2 instance enter below command:

• apt-get update
• apt-get install apache2
• /etc/init.d/apache2 restart

To edit the index.html file you will need to edit it by entering command “vim /var/www/index.html”. To enter the Insert Mode, press the “I” key alphabet on the keyboard. To save press “esc” then “:” + “w” + “q” and hit “Enter” after that it will exit you from the editor.

xix. When you are done, visit the http://the_public_IP_address_of_your_instance/ in the web browser. You should see IT works!

xx. The EC2 instance is still up running. It has been up since we have been logged into the AWS Virtual Machine. In order to install the Apache Server and edit the index.html file we have logout of the Virtual Machine and redirect to the NKU Virtual Machine in order to create S3 bucket type “exit” to do it. It will help you to logout of the root user access account of AWS. To achieve the project objective you will need to and configure S3 storage which can be done by following the command as seen in below picture ( Be sure to include your username not username you see in below).

You will need to create directory named “s3www” go inside that directory and create a html file called “index.html” using the vim. When you edit the file you will need the following content to be included in the file :-

It works! This website is hosted by Amazon AWS S3. When you finish the above step, you will need to upload the file and make it available in the bucket to the public by using following image command and you will see output as seen in below image:

To access the S3 object you will navigate to the browser and insert the link http://s3.amazonaws.com/your_user_name/index.html.

In order to start the static web hosting, we need to enable first the bucket and to do so you will use command “aws s3 website s3://your_user_name/ --index-document index.html”. When you enter the command in your web browser you will see message on the saying It works and below it will say “This website is hosted by Amazon AWS S3”.

Finally, we have S3 Bucket and EC2 instance been configured and are ready to perform the performance testing. In order to carry out the performance test, the Apache HTTP Server benchmarking tool which is also called ab will be used to conduct the performance test. It will display the results on how many requests per second the Apache installation can be served.

• ab -c 50 -n 5000 http://your_EC2_public_ip/
• ab -c 100 -n 5000 http://your_EC2_public_ip/
• ab -c 200 -n 5000 http://your_EC2_public_ip/
• ab -c 400 -n 5000 http://your_EC2_public_ip/
• ab -c 800 -n 5000 http://your_EC2_public_ip/
• ab -c 50 -n 5000 http://your_username.s3-website-us-east-1.amazonaws.com/
• ab -c 100 -n 5000 http:// your_username.s3-website-us-east-1.amazonaws.com/
• ab -c 200 -n 5000 http:// your_username.s3-website-us-east-1.amazonaws.com/
• ab -c 400 -n 5000 http:// your_username.s3-website-us-east-1.amazonaws.com/
• ab -c 800 -n 5000 http:// your_username.s3-website-us-east-1.amazonaws.com/

[Please record the “TIME PER REQUEST” value. Record the first value for each induvial test]

Data Collected. In this part of the project we will see the data which was collected from the Apache HTTP Server benchmarking tool. The test results are as follows:

• Test Results on EC2 Instance
• Test Results on S3 Bucket

Performance Analysis

The syntax of the ab command we used in the test is ab -c (According to apache the -c means number of multiple requests to perform at a line) and -n (According to apache the -n means number of the requests to perform for the benchmarking session).

Test 1, the concurrency number is 50 and number of the requests requested are 5000. The test results on time per request with the EC2 instance (76.359ms) result is higher than the S3 (73.231) which is lower than the time per request. So, the test 1 indicates EC2 process requests faster than the S3.

Test 2, the concurrency number is 100 and number of the requests requested are 5000. The test results on time per request with the EC2 instance (87.388ms) is lower than the time per request in S3(116.407ms). So, the test 2 indicates that S3 is processing requests faster than the EC2.

Test 3, the concurrency number is 200 and number of the requests requested are 5000. The test results on time per request with EC2(158.327ms) is higher than the S3(104.552ms. So, the test 3 concludes that EC2 has faster processing speed than the S3.

Test 4, the concurrency number is 400 and number the requests requested are 5000. The test results on time per request with EC2(269.689ms) is higher than the S3(159.6687). So, this test indicates that EC2 process the request of the clients faster than the S3.

Test 5, the concurrency number is 800 and number the requests requested are 5000. The test results on time per request with EC2(768.879ms) is higher than the S3(491.722ms). So, this test indicates that EC2 process the request of the clients faster than the S3.

Economic Analysis Data

In this part of the project the AWS services economic analysis is conducted. The AWS services are been charged on monthly basis like pay-as-you-go. The EC2 machine which I am going to discuss here is been configured with 1vCPU, 0.6 Memory and the service will run for entire 24 hours per day. The price of S3 for 100GB storage, 10000 PUT/COPY/POST/LIST requests, 10000 GET/SELECT and other requests, 1000GB data returned by S3 select and 1000GB data scanned by S3 Select.

As mentioned in above clustered graph you can clearly see the price comparison that the S3 is three times cheaper than the EC2 service. For better business cost cutting I would recommend them to use S3 service because it is easy for hosting the static website. Business would need to pay higher price for the EC2 service which can increase cost in the organization. S3 can perform best with the high ratio numbers of concurrency.

To save the cost for the static web hosting its much cheaper to adopt the S3 rather than EC2. To see better performance during high concurrency numbers the S3 gives outstanding responses than the EC2. S3 can be concluded to be better for the static web hosting because of its cost and performance.