CSE546 - Cloud Computing - Project 1

Created By:

Dhyey Pandya
Jaykumar Rajeshbhai Tandel
Sumedh Shridhar Joshi

Member Tasks

Dhyey Pandya:

I have been responsible for designing a part of the App Tier, which
involved the classification of the images received from AWS SQS
RequestQueue, followed by uploading the images to AWS S3 bucket called
‘cloud-project-1-input-images’ and the text files containing the image
name and its top-1 classification result to the bucket called
‘cloud-project-1-output-image-class’. I have also developed the App
Tier part to send the obtained classification result to the AWS SQS
ResponseQueue. I have also been involved with setting up the AWS SQS
Queues, S3 buckets and final testing of the entire application.

Jaykumar Rajeshbhai Tandel:

I designed the Flask web-app on an EC2 instance. I am responsible for
its functionalities consisting of accepting parallel requests coming
from the client, sending input images to the RequestQueue, accepting
responses coming from the ResponseQueue and sending correct responses
to the client.

Sumedh Shridhar Joshi:

I have been involved with designing the app-tier Amazon Machine Image
(AMI), setting up the Cloudwatch Scale-in and Scale-out alarms, and the
EC2 AutoScaling Group that would be in accordance with the Cloudwatch
alarms to add and remove the instances. I have been involved with
testing of the entire application and monitoring the status of the
alarms and AutoScale Instances.

Execution Steps and Services Details:

AWS Account Credntials:

Account ID: 104949213169
User: grader_user
Password: CloudGraderUser@3455

Web Tier:

Web Tier Name: web_tier
Web Tier URL: Here
Web Tier URL Once the Instance is Running: {web-tier-ip-address}

SQS:

Request Queue Name: RequestQueue
Response Queue Name: ResponseQueue

S3:

Input Images Bucket: cloud-project-1-input-images
Output Image/Classname Bucket: cloud-project-1-output-image-class

Cloudwatch Alarms:

Scale In Alarm Name: Scale in
Scale Out Alarm Name: Scale out

Steps To Execute:

1. Sign in to AWS as IAM user and enter the credentials given above.
2. Go to the web tier instance and start the instance.
3. After it starts running, go to the Connect section and follow the commands in SSH tab in the local terminal.
4. After connecting to the web tier instance, run the following commands:
   cd web_tier
   python3 web_tier.py
5. Now that the host is running, run the multithread_workload_generator.py file from the local machine using the web tier's public IP address in the --url argument. For example:
   python3 multithread_workload_generator.py
   --num_request 100
   --url "http://{web_tier_public_ip}:5000/accept_images"
   --image_folder "/path/to/the/image-net/folder/imagenet-100/"
6. You can check the CLoudwatch Alarms, EC2 Instances, SQS Queues, and S3 Buckets to monitor the progress in the respective services.
7. After testing, go to the web tier instance, and stop it.