Default region name: us-east-1

Default output format: json

iii.      After configuration, a folder named .aws gets created in the user ’s home directory.

iv.       Inside the .aws directory, you will see the following two files.

1. config: It contains all the default configs like region and output. You can change these values anytime and add new values as default.

2. credentials: This file contains the access key and secret key as plaintext.

5. Creating an AWS Project (Code examples shown with Python)

a. Creating EC2 Instances: The following steps to create EC2 instances:

i.       Define your security credentials

1.   Create your Key pairs in the EC2 dashboard

2.   Copy the pem file into your ~/.aws folder

3.   Add  SSH  Inbound  rules  so  that  you  can  log  in  to  your  instance through an SSH session (You may need to add more rules for your project)

4.   Code example to define security credentials

AWS_ACCESS_KEY_ID     = "YOUR AccessKeyId"

AWS_SECRET_ACCESS_KEY = "YOUR AWSSecret Key"

ii. Setup an Amazon EC2 Client:

1. Use the region of “US_EAST_1”

ec2 = boto3.resource(

'ec2',

region_name='US_EAST_1',

aws_access_key_id=config('AWS_ACCESS_KEY_ID'),

aws_secret_access_key=config('AWS_SECRET_ACCESS_KEY'))

iii.       Launch an EC2 instance

1.   Use AMI “ami-00ddb0e5626798373” for basic Ubuntu 18.04, as shown in this example.

ami_id = "ami-066ea1555baeb8dc4"

instance = ec2_client.create_instances(

ImageId=ami_id,

MinCount=1,

MaxCount=1,

InstanceType="t2.micro",

TagSpecifications=[{'ResourceType':'instance',

'Tags': [{

'Key': 'Name',

'Value': 'WebTier' }]}])

iv.       Set tags for the EC2 Instance

CreateTagsRequest createTagsRequest = new

CreateTagsRequest().withResources(

instance.getInstanceId())

.withTags(new Tag("Name", "Your Tag Name"));

ec2Client.createTags(createTagsRequest);

v. Start the Instance

instance = ec2_client.run_instances(

ImageId=ami_id,

MinCount=1,

MaxCount=1,

InstanceType="t2.micro",

TagSpecifications=[{'ResourceType': 'instance',

'Tags': [{

'Key': 'Name',

'Value': 'WebTier Worker' }]}])

vi. Check the state of the instance.

You can see that the created EC2 instance is running

vii. Connect to the Instance:

You should be able to loginto your instance with ssh session

STEP-2: Developing the Web Tier

The web tier will receive face recognition requests from clients and return the classification    results to the clients. In Part 1, instead of using a real classification model, which requires the app tier and data tier, we will use a lookup table to emulate the model inference process.

We will develop the web tier using a single EC2 micro instance. Assign this instance a static IP address so it does not change during development and testing.

An  Elastic  IP  address  is  a  static   IPv4  address  designed  for  dynamic  cloud  computing.  It  is allocated  to  your  AWS  account  and  yours  until  you  release  it.   Follow  the  following   help documents to determine how to assign an Elastic IP address to an EC2 Instance.

● Elastic IP addresses - Amazon Elastic Compute Cloud

● How to Add a Static IP to an AWS EC2 Instance - DEV Community

● allocate_address - Boto3 1.34.27 documentation (amazonaws.com)

The web tier should fulfill the below requirements:

1.   It  should  take  images  received  from  users  as  input  and  perform  face  recognition  on these  images  by  looking  up  the  classification  results  provided  to  you.  It  should  also return the recognition result as output to the users. The input from each request is a .jpg file, and the output in each response is the prediction result.  Find more details below.

Input:

●   The key to the HTTP payload MUST be defined as “inputFile” and should be used as the same. In the case of a Python backend, it denotes a standard Python file object.

● For example, the user uploads an image named “test_00.jpg ”.

●   Use the provided workload generator to generate requests to your web tier.

Output:

● The web tier will handle HTTP POST requests to the root endpoint ("/").

●   The output MUST be in plaintext, and the format :

● For the above example request, the output should be “test_00:Paul” in plaintext.

● You need to implement the handling of concurrent requests in your web tier.

To facilitate the testing, a standard face dataset and the expected recognition output of each image are provided to you at:

● Input: visa-lab/CSE546-Cloud-Computing/face_images_1000.zip

● Output: visa-lab/CSE546-Cloud-Computing/classification_face_images_1000.csv

2.   The  web tier should be able to handle multiple requests concurrently and as quickly as possible. The recognition results should all be correct. For 100 concurrent requests, the total runtime is about 13.67 seconds; for 1000 concurrent requests, the total runtime is about 2.04 seconds.

3.   To  facilitate testing, you MUST use only the resources from the US-East-1 region, and you MUST name your web-tier instance “web-instance”