Abstraction
Simplifying the Complex by Hiding Complicated Details
Abstraction sounds like a scary, complex word when you encounter it for the first time, but it is a universal concept that transcends boundaries across various industries.
From the precise realm of mathematics to the creative expanse of the arts. It’s a powerful tool that enables us to simplify complex realities, focus on essential features, and convey intricate ideas in a more digestible form. By omitting unnecessary details and emphasizing key elements, abstraction helps us comprehend, communicate, and innovate more effectively.
Abstraction as a Concept Across Disciplines
Since abstraction is not exclusive to the cloud, these definitions and analogies demonstrate its far-reaching applicability across various fields, allowing us to develop a more sophisticated and nuanced comprehension of the concept.
Abstraction in Computer Science
Abstraction is the process of exposing essential features while hiding implementation details, enabling modular programming and software development.
In Computer Science, Abstraction is like building a LEGO castle. You create a structure (program) using individual blocks (code) that hide the details of how they’re assembled, allowing you to focus on the overall design.
Abstraction in Engineering
Abstraction is used to design and develop systems, focusing on functional requirements while hiding implementation details.
Abstraction is like designing a car. You focus on functional requirements (speed, safety) while hiding implementation details (engine, transmission).
Abstraction in Mathematics
Abstraction involves generalizing specific instances into broader concepts, like grouping numbers into sets or defining algebraic structures.
Abstraction is like grouping different types of fruit (apples, bananas, oranges) into a single category (fruit). You focus on shared characteristics, ignoring specific details.
Abstraction in Art
Abstraction refers to non-representational art that doesn’t depict realistic subjects, instead focusing on colors, shapes, and forms to convey emotions and ideas.
Abstraction is like creating a puzzle. You take individual pieces (colors, shapes) and combine them to form a larger image that conveys a message or emotion, rather than a literal representation.
Abstraction in Philosophy
Abstraction is the mental process of isolating essential properties or characteristics from specific instances, leading to general concepts and theories.
Abstraction is like looking at a forest instead of individual trees. You step back to see the bigger picture.
Abstraction in Science
Abstraction entails simplifying complex systems or phenomena into theoretical models or representations, facilitating understanding and analysis.
Abstraction is like creating a map. You simplify complex terrain (data) into a representative model (theory), highlighting key features and relationships.
Abstraction in Music
Abstraction in music refers to experimental or avant-garde compositions that depart from traditional melodies and harmonies, emphasizing sound textures and atmospheres.
Abstraction is like creating a soundscape. You combine individual sounds (instruments, effects) to evoke an atmosphere or mood, rather than a specific melody.
Abstraction in Literature
Abstraction in literature involves using symbolic language, metaphors, and allegories to convey complex ideas and themes beyond literal interpretations.
Abstraction is like writing a poem. You use words and metaphors to convey emotions and ideas, rather than literal descriptions.
Abstraction in Economics
Abstraction in economics involves modeling complex systems, simplifying assumptions, and analyzing economic phenomena through theoretical frameworks.
Abstraction is like creating a financial model. You simplify complex systems (markets, economies) into theoretical frameworks to analyze and predict behavior.
Abstraction in Cognitive Science
Abstraction is the cognitive process of categorizing, generalizing, and forming mental representations of the world.
Abstraction is like categorizing animals into species. You group individual instances (animals) based on shared characteristics, forming mental representations.
Abstraction in Linguistics
Abstraction in linguistics involves analyzing language structures, identifying patterns, and developing theoretical models of language acquisition and use.
Abstraction is like learning a language. You identify patterns and structures (grammar, syntax) to understand and communicate effectively.
Abstraction in Business
Abstraction in business refers to identifying essential factors, simplifying complex processes, and developing strategic models to drive decision-making.
Abstraction is like developing a marketing strategy. You identify essential factors (target audience, messaging) and simplify complex processes to drive decision-making.
Abstraction in Psychology
Abstraction is the mental process of forming concepts, categorizing experiences, and developing mental representations of the world.
Abstraction is like forming memories. You categorize experiences and emotions, creating mental representations that influence future behaviors.
Abstraction in Architecture
Abstraction in architecture involves designing buildings and spaces that prioritize functional and aesthetic essentials over ornamental details.
Abstraction is like designing a building facade. You prioritize functional and aesthetic essentials (structure, appearance) over ornamental details.
Abstraction in Data Analysis
Abstraction in data analysis entails identifying key variables, simplifying complex data sets, and developing models to extract insights.
Abstraction is like creating a dashboard. You identify key variables (metrics, trends) and simplify complex data sets to extract insights and inform decisions.
…You get the point. If you’re transitioning to the cloud from one of the aforementioned industries, you can leverage your existing knowledge to understand what abstraction in the cloud means.
Abstraction in the Cloud
The Cloud builds upon the principles of computer science, and abstraction is a key concept that makes it more accessible. Think of abstraction like a filter that simplifies complex details by hiding them away from you, letting you focus on what matters.
In programming, abstraction helps hide intricate code details, allowing developers to concentrate on the logic of their application. Similarly, cloud providers use abstraction to conceal the complexity of infrastructure, offering pre-built services and managed platforms that are easy to use.
This abstraction enables you to focus on your goals, like deploying and managing applications, without worrying about the underlying infrastructure. It’s like having a trusted partner handle the complicated details, so you can concentrate on what you do best.
Putting it All Together with An Analogy
Abstraction in cloud computing is like the difference between driving a manual car versus driving an automatic vehicle, where a manual car is traditional on-premises infrastructure, and the automatic vehicle is the cloud.
Manual Cars (On Premises)
No Abstraction
Driving a manual (stick-shift) car requires direct control over gears and the clutch: In manual cars, you have fine-grained control but must manage every detail (shifting gears, clutch coordination). Similarly, on-premise solutions involve managing infrastructure at a granular level (physical servers, networking, security).
Automatic Cars (the Cloud)
An Abstraction of Manual Cars
Automatic cars handle gear changes automatically, simplifying driving:
Cloud services abstract away low-level infrastructure tasks. Just as automatic cars free you from gear management, cloud abstracts infrastructure complexities.
So would you rather be shifting gears the whole day, dealing with the clutch, waiting eons for the car to start in cold weather, and likely cause unnecessary accidents when ascending uphill on a rainy day? (On-Premises)
or
Do you want to just push a start button and go, where you just steer yourself in the right direction, accelerate when you need to and easily stop when you want? (Cloud Computing).
This analogy emphasizes that the core complicated, technical components are not completely gone, they are simply hidden and managed by the cloud provider, simplifying your experience so you innovate and build faster.
Let’s Build: Demonstrating Abstraction in the Cloud
I know (hope) you’re eager to set something up, so in this practical, hands-on-tutorial, we will deploy a basic static site. This will demonstrate abstraction by contrasting Amazon Elastic Compute Cloud (EC2) and Amazon Beanstalk, which is considered an abstraction of EC2.
Prerequisites
Add Additional Permissions to Your IAM Account
Principle of Least Privilege: It is considered a best practice to grant users only the necessary permissions to perform their tasks, it is a fundamental security concept that will help keep your systems secure in the long run.
Step 1: Sign in with your Administrative Root User Account in order to grant permissions to your active Alias or IAM Account.
Step 2: Select Security, Identity & Compliance on the left side of the drop down menu.
Step 3: Select IAM.
You should be redirected to the IAM Dashboard.
Step 4: On the left side of the screen expand Access management.
Step 5: Select Users.
Step 6: Select the IAM User.
Step 7: Select Permissions.
Click Add permissions.
Select Add Permissions.
You should be redirected to the Add permissions page.
Step 8: Select Add policies directly.
Step 9: On the Search bar type and check: AmazonEC2FullAccess.
On the Search bar type and check: AdministratorAccess-AWSElasticBeanstalk.
On the Search bar type and check: AmazonS3FullAccess
On the Search bar type and check: AWSCloudFormationFullAccess
On the Search bar type and check: AmazonRDSFullAccess.
On the Search bar type and check: CloudWatchFullAccess
Step 10: Click Next.
You should be redirected to the Review page.
Step 11: Click Add permissions.
Step 12: If you performed the steps successfully, you should see a green banner: 6 policies added.
Create an IAM EC2 Instance Profile for Elastic Beanstalk
Step 13: Below Users. Select Roles.
Step 14: Click Create role.
Step 15: You should be redirected to the Select trusted entity page.
Select AWS service.
Use Case: EC2.
Click Next.
Step 16: Add permissions. Select AdministratorAccessAWSElasticBeanstalk
Click Next.
You should be redirected to the Name, review, and create page.
Step 17: Role name: AbstractionRole
Click Create role.
Step 18: If you performed the steps successfully, you should see a green banner.
Click View role.
Build Using Your IAM Account
Step 19: Log out of your Administrative Account, and Sign in with your IAM Account.
Depending on how long ago you last logged in and your password policy, you might have to reset your password using your Administrative Account. Ignore Steps 20.1 —20.8 if you don’t have to reset your password.
Step 20.1: Sign in with your Administrative Root User Account in order to reset your password.
Step 20.2: Select Security, Identity & Compliance on the left side of the drop down menu.
Step 20.3: Select IAM.
You should be redirected to the IAM Dashboard.
Step 20.4: On the left side of the screen expand Access management.
Step 20.5: Select Users.
Step 20.6: Select the IAM User.
Step 20.7: Select Security credentials.
Click Manage console access.
Select Reset password.
Under Console password, Select Custom password.
Click Reset password.
If you performed the steps successfully, you should see a green banner:
You have successfully enabled the user’s new password.
This is the only time you can view this password. After you close this window, if the password is lost, you must create a new one.
Step 20.8: Click Close.
Elastic Compute Cloud Tutorial
Before you proceed, make sure that you’re signed in with your IAM User Account, and that you’re in the N.Virginia region (us-east-1).
Launch EC2 Instances
Step 21: Select Compute on the left side of the drop down menu.
Step 22: Select EC2.
Step 23: Click Launch instance.
Step 24: Under Name and Tags, Name your instance NoAbstraction.
Step 25: Under Application and OS Images (Amazon Machine Image).
Click Quick Start.
Select Amazon Linux.
Under Amazon Machine Image (AMI), select Amazon Linux 2 AMI (HVM) (Free Tier eligible).
Under Architecture, select 64-bit (x86).
Step 26: Under Instance type, select t2.micro (Free tier eligible).
Step 27: Under Key pair (login), select Create new key pair.
You should be redirected to the Create key pair pop up.
Key pair name: AbstractionKey.
Key pair type: RSA
Private key file format: .pem
Click Create key pair.
Save it on your local device.
Step 28: Under Network settings.
Click Edit.
Navigate to Firewall (security groups).
Select Create security group.
Security group name — required: type in MySecurityGroup.
Description — required: type in Security Group for My Web Servers.
Under Inbound Security Group Rules.
Click Add Security group rule.
Type: Select HTTP.
Source Type: Select Anywhere.
Step 29: Configure storage: leave as default.
Step 30: Advanced details: leave as default, scroll to the bottom, but, don’t scroll too fast, make a note of all the features you can manually configure until you reach User data — optional.
Paste this:
#!/bin/bash
# Update and install dependencies
sudo yum update -y
sudo yum install -y python3
sudo pip3 install flask
# Create a directory for the app
mkdir /home/ec2-user/hello-world
# Create the Flask application
cat > /home/ec2-user/hello-world/app.py <<EOF
from flask import Flask
app = Flask(__name__)
@app.route('/')
def hello():
return "Hello World from EC2! This is NO ABSTRACTION"
if __name__ == '__main__':
app.run(host='0.0.0.0', port=80)
EOF
# Run the Flask app
nohup python3 /home/ec2-user/hello-world/app.py &Step 31: At the right of the page Summary.
Number of instances: 2.
Click Launch instance.
If you performed the steps successfully, you should see a green banner:
Success.
Successfully initiated launch of instances.
Step 32: At the bottom of the page, click View all instances.
You should be redirected to the instances dashboard. Make sure your instances are in the Running state.
Step 33: Click on the Instance ID of the first instance.
Step 34: Copy the Public IPv4 address, and paste it into your browser.
Note: Make sure it begins with http:// (Port 80) and not https:// (Port 443). You will received ERR_CONNECTION_TIMED_OUT if it begins with https:// because the Security group associated with your instance is open on Port 80 (HTTP) and not Port 443 (HTTPS).
Step 35: Congratulations! If you performed the steps successfully, you should see the following message:
Hello World from EC2! This is NO ABSTRACTION
Configure AutoScaling
Autoscaling is like having extra help ready to step in when needed.
In simple terms, autoscaling is a feature that automatically adjusts the number of servers (instances) your application uses based on how much traffic or work it’s handling.
If your app gets busy (like during a sale or a viral moment), autoscaling adds more servers to handle the load.
If things get quiet, autoscaling removes extra servers to save money.
This way, your application always has just the right amount of resources — never too many (wasting money) or too few (causing slowdowns or crashes).
Step 36: Navigate back to the Console. On the far right of the page, at the bottom, below Autoscaling, select Auto Scaling Groups.
Step 37: Click Create Auto Scaling Group.
You should be redirected to the Choose launch template or configuration page.
Name. Auto Scaling group name: AutoScalingGroup.
Launch template: Launch template: Click Create a launch template
You should be redirected to the Create launch template page.
Create launch template
Launch template name and description.
Launch template name — required: MyLaunchTemplate.
Template version description: Version 1
Auto Scaling guidance: Unchecked.
Launch template contents
Application and OS Images (Amazon Machine Image)
Select Currently in use.
Instance Type: t3.micro.
Key pair (login): AbstractionKey.
Network Settings
Subnet: Don’t include in launch template.
Firewall (Security groups): Select existing security group.
Common security groups: MySecurityGroup.
Storage (volumes): leave default.
Resource tags: leave default.
Advanced details: Scroll to the bottom and paste the code below.
#!/bin/bash
# Update and install dependencies
sudo yum update -y
sudo yum install -y python3
sudo pip3 install flask
# Create a directory for the app
mkdir /home/ec2-user/hello-world
# Create the Flask application
cat > /home/ec2-user/hello-world/app.py <<EOF
from flask import Flask
app = Flask(__name__)
@app.route('/')
def hello():
return "Autoscaling: This is NO ABSTRACTION"
if __name__ == '__main__':
app.run(host='0.0.0.0', port=80)
EOF
# Run the Flask app
nohup python3 /home/ec2-user/hello-world/app.py &Summary: click Create launch template.
If you performed the steps successfully, you should see a green banner:
Success.
Successfully created MyLaunchTemplate.
Click View launch templates.
Navigate back to the Choose launch template or configuration page.
Click the refresh button next to Select a launch template.
Select MyLaunchTemplate.
Click Next.
Step 38: Network: ensure defaultVPC is selected.
Availability Zones and subnets: select all.
Availability Zone distribution — new: select Balanced best effort.
Click Next.
Load balancing
Load balancing is like having a traffic cop for your application.
When lots of users try to access your app at the same time, a load balancer makes sure the traffic is spread out evenly across all your servers. This prevents any single server from being overwhelmed while others sit idle.
It ensures:
Your app runs smoothly, even under heavy traffic.
If one server fails, the load balancer sends traffic to the healthy servers so users don’t notice any issues.
Think of it as distributing the workload to keep everything running efficiently!
Attach to a new load balancer
Step 39: Load balancing: select Attach to a new loadbalancer.
Load balancer type: select Application Load Balancer (HTTP, HTTPS)
Load balancer name: MyLoadBalancer.
Load balancer scheme: Internet-facing.
Network mapping: Leave default.
Listeners and routing:
Protocol: HTTP
Port: 80
Default routing (forward to): Create a target group.
New target group name: MyTargetGroup.
VPC Lattice integration options: No VPC Lattic service.
Leave everything as default.
Click Next.
Step 40: Configure group size and scaling — optional.
Desired capacity: 2.
Scaling.
Min desired capacity: 1
Max desired capacity: 2
Automatic scaling — optional: No scaling policies.
Instance maintenance policy: No policy.
Additional capacity settings: Default.
Additional settings: Leave all unchecked.
Click Next.
Step 41: Add notifications — optional: click Next.
Step 42: Add tags — optional: click Next.
Step 43: Review: Click Create Auto Scaling Group.
If you performed the steps successfully, you should be looking at the Auto Scaling groups page.
Navigate back to your instances page.
You should have three running instances.
Step 44: Check the instance we just created from the Auto Scaling Group.
Instance state: Terminate (delete) instance.
Terminate (delete) instance? : click Terminate (delete)
As soon as the instance is Terminated, your Auto Scaling group will automatically launch a new instance.
When this instance is in running state, click on the Instance ID of the newly launched instance.
Copy the Public IPv4 address, and paste it into your browser.
Note: Make sure it begins with http:// (Port 80) and not https:// (Port 443). You will received ERR_CONNECTION_TIMED_OUT if it begins with https:// because the Security group associated with your instance is open on Port 80 (HTTP) and not Port 443 (HTTPS).
If you performed the steps successfully, you should see the following message:
Autoscaling: This is NO ABSTRACTION
Step 45: Navigate to Load Balancing.
Select Target Groups.
Click MyTargetGroup.
Targets.
Registered Targets.
Click Register Targets.
Select all Available Instances.
Click Include as pending below.
Click Register pending targets.
Wait for all of them to reach Healthy status.
Step 46: Navigate to Load Balancing.
Click Load Balancers.
Click MyLoadBalancer.
Copy the DNS name and paste it into your browse, you should see (Instance One):
Hello World from EC2: This is NO ABSTRACTION
Reload the Web Page, you should see (Instance Two):
Hello World from EC2: This is NO ABSTRACTION
Reload the Web Page, you should see (Instance Three):
Autoscaling: This is NO ABSTRACTION
You might not see it in this particular order, but your Load Balancer will attempt to distribute traffic equally from all the instances you registered as targets.
Elastic Beanstalk Tutorial
Elastic Beanstalk is a service that makes it easy to run your application in the cloud without worrying about the servers.
You just upload your code, and Beanstalk automatically sets up everything you need, like servers (EC2), load balancers, scaling, and updates. It handles the heavy lifting so you can focus on your app, not the infrastructure.
Step 47: Navigate back to the Amazon Management Console.
Step 48: Click On Services.
Step 49: Select Compute.
Step 50: Click on Elastic Beanstalk.
Step 51: Under Get Started, click Create Application.
You should be redirected to the Configure environment page.
Step 52: Environment Tier: select Web server environment.
Application information. Application name: type Abstraction.
Environment name: leave default.
Domain: choose available domain name.
Environment description: type This is Abstraction.
Platform. Platform type. Select Managed platform.
Platform. Choose a platform: Python
Platform branch: Python 3.12 running on 64bit Amazon Linux 2023.
Platform version: 4.3.1
Application Code: Sample application
Presets: Single instance (free tier eligible)
Click Next.
You should be redirected to the Configure server access page.
Step 53: Service Role: Create and use new service role
Service role name: aws-elasticbeanstalk-service-role
EC2 key pair: AbstractionKey
EC2 instance profile: AbstractionRole.
Click Next.
You should be redirected to the Set up networking, database, and tags — optional page.
Step 54:
Virtual Private Cloud
Select defaultVPC.
Instance settings
Public IP address. Check Activated.
Instance Subnets: Check all.
Database
Choose database subnets: Select all.
Enable database,
Restore a snapshot — optional
Snapshot: None.
Database settings: Leave default excpet for Username and Password
Username: admin
Password: AdminPassword#2024
Database deletion policy: Delete.
Click Next.
You should be redirected to the Configure instance traffic and scaling — optional page.
Step 55:
Instances
Root volume (boot device). Root volume type: General Purpose 3 (SSD)
Scroll to EC2 Security Groups. Check MySecurityGroup.
Capacity
Auto scaling group.
Environment type: Load balanced.
Instances
2 Min
4 Max
Fleet composition: On-Demand Instances.
Architecture: x86_64
Instance types. Select t2.micro
Leave the rest as default.
Step 56:
Load balancer network settings
Load balancer subnets: select All.
Load Balancer Type:
Select Application Load Balancer.
Select Dedicated.
Listeners: 80
Processes: default.
Leave the rest as default.
Click Next.
You should be redirected to the Configure updates, monitoring, and logging — optional page.
Step 57:
Monitoring
Health reporting: Basic.
Managed platform updates
Activated: Uncheck.
Leave the rest as default.
Click Next.
Step 58: Click Submit.
Blue Banner: Elastic Beanstalk is launching your environment. This will take a few minutes.
Observe Elastic Beanstalk in action as it spins up your resources.
If you performed the steps successfully, you should eventually see a green banner:
Environment successfully launched.
And your Health status should move from Grey to Green.
Step 59: Click on the Domain.
Congratulations
Your first AWS Elastic Beanstalk Python Application is now running on your own dedicated environment in the AWS Cloud
This environment is launched with Elastic Beanstalk Python Platform
If you want to see the Instances launched by Beanstalk, you can head over to your Instances page.
Step 60: Click on the Abstraction instance.
Copy the Public IPv4 address, and paste it into your browser.
Note: Make sure it begins with http:// (Port 80) and not https:// (Port 443). You will received ERR_CONNECTION_TIMED_OUT if it begins with https:// because the Security group associated with your instance is open on Port 80 (HTTP) and not Port 443 (HTTPS).
You should see the same:
Congratulations
Your first AWS Elastic Beanstalk Python Application is now running on your own dedicated environment in the AWS Cloud
This environment is launched with Elastic Beanstalk Python Platform
Important: Clean Up Resources
Don’t get a Bill Shock by leaving unnecessary resources running.
- Navigate to Elastic Beanstalk and Delete your Application.
- Navigate to EC2 and Terminate your Instances.
- Navigate to Security Groups and Delete your Security Groups.
- Navigate to Auto Scaling Groups and Delete your Launch Template.
- Navigate to Load balancers and Delete your Load Balancer.
- Navigate Target Groups and Delete your Target Group.
- Login with your Root User Account, and Remove All Permissions Associated with your IAM Account.
End of Tutorial.
Building Tutorial Overview
We observed how Elastic Beanstalk is an abstraction of EC2. With EC2, you are responsible for setting up everything yourself: launching servers, installing software like Python, and managing scaling, load balancing, and updates.
In contrast, Elastic Beanstalk simplified this process by letting you upload your code, and it automatically handled the underlying tasks for you.
Behind the scenes, Beanstalk still uses EC2 instances, but it abstracts away the complexity, so you can focus on building and running your application instead of managing infrastructure.
Summary
To summarize, when you hear or see abstraction or ‘x is an abstraction of y’ in the cloud, it simply means that the cloud provider has taken care of the intricate details, allowing you to work more efficiently and productively.
Abstraction makes cloud computing more accessible and user-friendly, helping you achieve your goals without getting bogged down in technical complexities.
Prerequisites:
You want to understand abstraction in the cloud deeply.
Theory:
- Abstraction
- Abstraction as a Concept Across Disciplines
- Abstraction in the Cloud
- Putting It All Together with an Analogy: Manual vs Automatic Cars
Hands-On:
- Add Additional Permissions to Your IAM Account
- Create an IAM EC2 Instance Profile for Elastic Beanstalk
- Build Using Your IAM Account
- Reset IAM User Password
- Launch EC2 Instances
- Configure Auto Scaling
- Setup Load Balancing
- Create Sample Application with Elastic Beanstalk
Additional Resources
What is Amazon EC2?
Amazon Elastic Compute Cloud (Amazon EC2) is a web service that provides secure, resizable compute capacity in the cloud.
Why Elastic Beanstalk?
Elastic Beanstalk is a service for deploying and scaling web applications and services. Upload your code and Elastic Beanstalk automatically handles the deployment — from capacity provisioning, load balancing, and auto scaling to application health monitoring
Abstractions: Creating the best developer experience
Concluding Remarks
You will encounter the term abstraction frequently when reading documentation about cloud services, especially when it comes to newer, evolved releases. Your task is not necessarily to take action, but rather to understand and appreciate what the cloud provider has abstracted away to simplify certain services for your benefit.
