Version: 3.11.2022 v2.1

Summary

This coursework is divided into two parts:

Part 1: A written task (only) related to the knowledge gained in the AWS Academy Cloud Foundations course (weeks 1-7).

Part 2: A combined practical and written activity architecting a scaling application on the Cloud,         where you will be required to use knowledge gained and a little further research to implement the   scaling infrastructure, followed by a report that will focus on your experience in the practical activity together with knowledge gained in the entire LSDE course.

Weighting: This assessment is worth 100% of your total unit 20 credits.

Set: 13:00. Monday 28th  Nov 2022.

Due: 13:00. Tuesday 10th Jan 2023.

Pre-requisites:

•    You must have completed the AWS Academy Cloud Foundations course set in weeks 1-7

•    You will require an AWS Academy Learner Lab account for the practical activity. You should receive an invite when this document is released. Please contact the LSDE Unit Director if   you have no email or issues with the registration.

•    A Secure Shell (SSH) client, such as MacOS Terminal or PuTTy on Windows, for server admin.

Submission:

Via the LSDE BlackBoard coursework assessment page, submit one zip file, named using your UOB username (‘username.zip’), containing:

•    a Report (‘report.pdf’) in PDF format containing:

o Part 2

•    a Text File (‘credentials.txt’) containing your AWS Academy account credentials (username,

password), to enable us to access and review your Learner Lab account as required.     In this document we provide a detailed explanation of the tasks, and the approach to marking.

Unit Director: Jin Zheng

Task 1: (25%)

Write a maximum of 1000 words (minimum: 600) debating the statement:

" Financial institutions should use public cloud for data processing"

Include your own descriptions of the following:

•    At least 5 AWS features or services introduced in the Cloud Foundations course that make data processing in the public cloud advantageous for financial institutions.

•    At least 3 scenarios where the public cloud should be avoided for data processing for financial institutions.

Task 2: Scaling the WordFreq Application (75%)

Overview

WordFreq is a complete, working application, built using the Go programming language.

]NOTE: (on oJ? NOT ?xd?O??p ?o nnp?Js?onp oJ d?Jmi???p ?o mopiJ( ?V? sonJO? Oop? in on( wo([

The basic functionality of the application is to count words in a text file. It returns the top ten most frequent words found in a text document and can process multiple text files sequentially.

The application uses a number of AWS services:

•    S3: There are two S3 buckets used for the application.

o One is used for uploading and storing original text files from your local machine. This is your uploading bucket.

o These files will be copied from the uploading bucket to the processing S3 bucket. The bucket has upload notifications enabled, such that when a file is uploaded, a message notification is automatically added to a wordfreq SQS queue and an email will be sent to you.

•    SQS: There are two queues used for the application.

o One is used for holding notification messages of newly uploaded text files from the S3 bucket. These messages are known as ‘jobs’, or tasks to be performed by the application, and specify the location of the text file on the S3 bucket.

o A second queue is used to hold messages containing the ‘top 10’ results of the processed jobs.

•    SNS: Publishes messages to your email address and SQS queues.

•    DynamoDB: A NoSQL database table is created to store the results of the processed jobs.

•    EC2: The application runs on an Ubuntu Linux EC2 instance, which you will need to set up initially following the instructions given. This will include setting up and identifying the S3, SQS and DynamoDB resources to the application.

You will be required to initially set up and test the application, using instructions given with the zip download file. You will then need to implement auto-scaling for the application and improve its architecture based on principles learned in the CF course. Finally, you will write a report covering   this process, along with some extra material.

Figure 1 - WordFreq standard architecture

Task A – Install the Application

Ensure you have accepted access to your AWS Academy Learner Lab account and have at least $40 credit (you are provided with $100 to start with). If you are running short of credit, please inform your instructor.

Refer to the WordFreq installation instructions (‘README.txt’) in the coursework zip download on the BlackBoard site, to install and configure the application in your Learner Lab account. These instructions do not cover every step – you are assumed to be confident in certain tasks, such as in the use of IAM permissions, launching and connecting via SSH to an EC2 instance, etc.

You will set up the database, storage buckets, queues and worker EC2 instance.

Finally, ensure that you can upload a file using the ‘run_upload.sh’ script and can see the results logged from the running worker service, before moving on to the next task.

]NOTE: TVa oddliOo?ion Oopa is in ?Va 9o lon6no6a. Yon oJa NOT axdaO?ap ?o nnpaJs?onp oJ mopiJ( i?. An( Oopa OVon6as will qa i6noJap onp mo( losa moJ入s.[

Task B – Design and Implement Auto-scaling

Review the architecture of the existing application. Each job process takes a random time to complete between 10-20 seconds (artificially induced, but DO NOT modify the application source code!). To be able to process multiple uploaded files, we need to add scaling to the application.

This should initially function as follows:

•    When a given maximum performance metric threshold is exceeded, an identical worker instance is launched and begins to also process messages on the queues.

•    When a given minimum performance metric threshold is exceeded, the most recently launched worker instance is removed (terminated).

•   There must always be at least one worker instance available to process messages when the application architecture is 'live'.

Using the knowledge gained from the Cloud Foundations course, architect and implement auto-scaling functionality for the WordFreq application. Note that this will not be exactly the same as Lab 6 in Module 10, which is for a web application. You will not need a load balancer, and you will need to identify a different CloudWatch performance metric to use for the ‘scale out’ and ‘scale in’ rules. The 'Average CPU Utilization' metric used in Lab 6 is not necessarily the best choice for this application.

Task C - Perform Load Testing

Once you have set up your auto-scaling infrastructure, test that it works. The simplest method is to create around 40 large text files. Please make sure you’ve uploaded 40 files to your uploading S3    bucket.

You can ‘purge’ all files from your processing S3 bucket, then you could copy all the .txt files from you uploading S3 bucket to your processing S3 bucket.

•    Connect to one of your instances that in your Auto Scaling Group (via SSH connection).

•    Copy all the .txt file from your uploading S3 bucket (e.g., zj-wordfreq-nov22-uploading) to your processing S3 bucket (e.g., zj-wordfreq-nov22-processing) by running the following   command in your SSH terminal:

aws s3 cp s3://<name of your uploading bucket> s3://<name of your

processing bucket> --exclude "*" --include "*.txt" --recursive Please watch the following behaviours:

•    Watch the behaviour of your application to check the scale out (add instances) and scale in (remove instances) functionality works.

•   Take screenshots of your copied files, the SQS queue page showing message status, the Auto Scaling Group page showing instance status and the EC2 instance page showing launched /    terminated instances during this process.

•   Take a screenshot of the emails you’ve received from Amazon S3 Notification. Ideally you    are expected to get 40 emails. You only need to take a screenshot of one email to show the functionality of SNS.

•   Try to optimise the scaling operation, for example so that instances are launched quickly   when required and terminated soon (but not immediately) when not required. Note down settings you used and the fastest file processing time you achieved.

•   Try using a few different EC2 instance types – with more CPU power, memory, etc. Note down any changes in processing time.

•   Please delete all the .txt file in your processing S3 bucket after load testing.

•   [NOTE: The Learner Lab accounts officially only allow a maximum of 9 instances running in one region, including auto-scaling instances. Learner Lab accounts are Limited in which EC2 Types and AWS services they can use. This is explained in the Lab Readme file on the Lab page; section ‘Service usage and other restrictions’.]

Task D - Optimise the WordFreq Architecture

Using services, features and techniques learned from the Cloud Foundations course, improve the architecture in the following areas

a)   Create an Amazon SNS topic and configure the file deleted notification from S3 Bucket to SNS so you could get notification by emails upon deleting the .txt files of your processing S3 bucket. Take a screenshot of one email you’ve received. You may want to create a new SNS topic and configure a new event notification in your processing S3 bucket for this topic.

b)   Based on only AWS services and features learned from the Cloud Foundations course, describe how you could re-design the WordFreq application’s current cloud architecture (i.e. not changing the application’s functionality or code) to suit the following requirements:

•    Increase resilience and availability of the application against component failure.

•    Long-term backups of valuable data required.

•    Cost-effective and efficient application for occasional use. Processing does not need to be immediate.

•    Prevent unauthorised access.

Your description should ideally include diagrams and include the AWS services required together with a high-level explanation of features & configuration for each, related to the following categories:

•    Data processing performance of the application as appropriate.

•    Resilience and availability of the application against component failure.

•    Security of the application for data protection and to prevent unauthorised access.

•    Ensuring the application is as cost-effective as required.

c)    Attempt to apply as much of your design as possible to your Task C WordFreq application architecture in the Learner Lab. If you do not have permissions or access to a particular service or feature, mention this after your design description in part b) above.

[NOTE: Ensure that your WordFreq application’s auto-scaling is still functional when finished!]

Task E - Create the Final Report

Write a report of no more than 3500 words and 20 A4 pages (there is NO minimum), including:

•    A brief summary of how the application works (without any reference to the code functionality).

•    Your design process to architect the scaling behaviours (task B).

•    An overview of the testing and your results, including screenshots (task C).

•    Your architectural description (task D).

•    Details of any issues you had and whether you resolved them.

Add one final section of the report: Further Improvements

•    Based on services and frameworks covered in the full LSDE course, identify two alternative data processing applications that would be far more performant and robust for this processing task.

[Do not implement these ideas, just describe their advantages over WordFreq in a few paragraphs].

The report should be included within the zip file as a PDF. It does not need to follow any academic format, but you should use grammar and spelling checkers on it and make good use of paragraphs and sub-headings. Double-spacing is not required. Use diagrams where they make sense and include captions & references from the text.

[IMPORTANT: All text not originally created by you must be cited, leading to a final numbered reference section (based on e.g. the British Standard Numeric System) to avoid accusations of plagiarism.]

[IMPORTANT: Disable autoscaling at end of each lab session: – Desired capacity = 0 ; Minimum capacity = 0. This saves credit and avoids multiple instances from launching and terminating when starting / stopping a lab session]

AWS Academy Learner Lab

You are given an AWS Academy Learner Lab account for this coursework. Each account has $100 assigned to it, which is updated every 24 hours and displayed on the Academy Lab page.

To access the lab from AWS Academy, select Courses > AWS Academy Learner Lab > Modules > Learner Lab - Foundation Services. On this page click ‘Start Lab’ to start a new lab session, then the ‘AWS’ link to open the AWS Console once the button beside the link is green.

Please note:

•    Ensure you shut down (stop or terminate) EC2 instances when you are not using them. These will use the most credit in your account in this exercise. Note that the Learner Lab will stop running instances when a session ends, then restart them when a new session begins.

•    AWS Learner Lab accounts have only a limited subset of AWS services / features available to them, see the Readme file on the Lab page (Service usage and other restrictions).

•    If you have installed the AWS CLI on your PC and wish to access your Learner Lab account,    you will need the credentials (access key ID & secret access key) shown by pressing the AWS Details button on the Lab page. Note that these only remain valid for the current session.

•   If you have any issues with AWS Academy or the Learner Lab, please book an Office Hours  session or use the LSDE Discussion Forums to seek help FIRST, email the instructors if there is no other option.

Support

The normal options for support are available for you during term time, up until Week 13 (January 10th ):

•    Book Office Hours for tech questions/support in LSDE Teams site form.

o If available dates are not shown, view page in browser:

https://outlook.office365.com/owa/calendar/LSDEOfficeHours20222023@bristol.ac .uk/bookings/s/EERrRf89gEqjpGwE4fGQ7g2

o Office Hours: 15 min sessions with a member of staff or a TA.

•    LSDE BlackBoard Discussions Forum