DEV Community

Yunus O. KAREEM
Yunus O. KAREEM

Posted on Edited on

Before You Create an AWS Account, Spend 45 Minutes Learning These Security Tools

Two clients. Two compromised cloud environments. One lesson: Cloud security begins before deployment.

The Two Incidents That Changed How I Talk About AWS Security

A few years ago, if someone asked me what they needed to know before deploying an application on AWS, my first response would probably have focused on EC2, S3, RDS, VPC, and perhaps Route 53.

Today, my answer is different.

Before you deploy anything on AWS, spend at least 45 minutes understanding its security tools.

Why?

Because I've encountered two client situations involving compromised AWS environments where insufficient security knowledge during infrastructure setup was a concern.

These experiences reinforced something I believe every developer, startup founder, DevOps engineer, and cloud enthusiast should understand:

AWS gives you powerful infrastructure, but using AWS does not automatically make your application secure.

In both situations, the important question was not simply how the attackers gained access. It was also whether the right security controls, monitoring, and access restrictions had been considered from the beginning.

And this is where many cloud beginners make their first mistake.

They spend hours learning how to launch an EC2 instance, configure a database, or deploy a Docker container.

But how much time do they spend learning how to secure those resources?

Often, not enough.

The Problem: We Learn How to Deploy Before We Learn How to Defend

Imagine a developer building a new SaaS product.

They create an AWS account, launch an EC2 instance, configure an RDS database, deploy their application, and connect a domain.

Everything works.

The application is live.

The founder is happy.

But several important questions remain unanswered:

  • Who can access the AWS account?
  • Are root credentials protected with multi-factor authentication?
  • Are any access keys exposed or unnecessarily long-lived?
  • Are the servers and databases publicly accessible?
  • Is someone monitoring suspicious API activity?
  • Would anyone receive an alert if unusual activity occurred?
  • What happens if a credential is compromised?

A working application is not necessarily a secure application.

And the cost of discovering this distinction after a security incident can be significant.

The good news is that AWS provides native tools to help organisations establish a stronger security baseline.

You do not need to master every AWS security service before creating your first application.

But you should know the essentials.

Here are the tools and controls I recommend learning in your first 45 minutes.


Minute 0–5: Understand the AWS Shared Responsibility Model

Before learning any tool, understand one fundamental concept: the AWS Shared Responsibility Model.

AWS is responsible for security of the cloud, including the physical infrastructure, hardware, and underlying facilities.

Customers are responsible for security in the cloud, with responsibilities varying according to the services they use.

For example, when you deploy an application on Amazon EC2, AWS manages the physical infrastructure, but you are responsible for tasks such as operating system security, firewall configuration, application vulnerabilities, and identity permissions.

This is the first misconception every cloud beginner must overcome.

AWS secures its infrastructure. You must secure what you build on it.

Reference: AWS Shared Responsibility Model

Minute 5–12: AWS IAM — Your First Line of Defence

If I could recommend one AWS security service that every beginner must understand, it would be AWS Identity and Access Management (IAM).

IAM controls who can access AWS resources and what actions they are permitted to perform.

Think of IAM as the security gatekeeper of your cloud environment.

What you must know

1. Protect your root user.

The AWS root user has extensive privileges. Enable MFA, avoid using the root user for everyday operations, and do not create root access keys.

2. Use IAM Identity Center.

For human access, prefer IAM Identity Center with temporary credentials rather than creating long-lived access keys.

3. Apply least privilege.

A developer who only needs access to an S3 bucket should not automatically receive administrator access to the entire AWS account.

4. Use IAM roles for applications.

Applications running on EC2, Lambda, or ECS should use appropriately scoped IAM roles instead of hardcoded AWS credentials.

Practical example

Suppose your application only needs to read objects from one S3 bucket.

An IAM policy could look like this:

{
  "Version": "2012-10-17",
  "Statement": [
    {
      "Effect": "Allow",
      "Action": "s3:GetObject",
      "Resource": "arn:aws:s3:::my-app-bucket/*"
    }
  ]
}
Enter fullscreen mode Exit fullscreen mode

This is significantly safer than granting AmazonS3FullAccess when only read access is required.

Takeaway: Never grant more permissions than an identity actually needs.

Minute 12–18: Amazon VPC and Security Groups — Protect Your Network

One of the most dangerous assumptions in cloud computing is that a server is safe simply because it is running inside AWS.

Amazon Virtual Private Cloud (VPC) allows you to define a logically isolated network for your AWS resources.

Security groups act as stateful virtual firewalls controlling inbound and outbound traffic.

Common mistakes to avoid

  • Allowing SSH (port 22) from 0.0.0.0/0.
  • Exposing databases directly to the internet.
  • Opening unnecessary application ports.
  • Placing sensitive workloads in public subnets without a valid requirement.

For example, an RDS database used by a web application should generally be deployed privately, with its security group allowing connections only from the authorised application security group.

For EC2 administration, consider AWS Systems Manager Session Manager rather than exposing SSH to the internet.

Your six-minute exercise

Open the AWS VPC console and locate the Security Groups section.

Understand the difference between inbound and outbound rules, public and private subnets, and security group references.

Takeaway: Every publicly accessible service expands your attack surface.

Minute 18–23: AWS CloudTrail — Know Who Did What

Imagine discovering that someone has created an unfamiliar IAM user or changed a security group rule.

How would you investigate?

This is where AWS CloudTrail becomes essential.

CloudTrail records supported AWS API activity and helps you understand actions performed within your AWS environment.

It can help answer questions such as:

  • Who created a new IAM role?
  • When was a security group modified?
  • Which identity attempted an unauthorised API operation?
  • Where did an API request originate?

CloudTrail Event History provides recent management-event visibility, but a properly configured trail or CloudTrail Lake event data store is needed for broader, longer-term auditing.

What to configure

Create a multi-Region CloudTrail trail, send logs to a protected S3 bucket, enable log file validation, and apply appropriate retention and access policies.

For sensitive environments, consider organisation-level logging in a dedicated security account.

Takeaway: If you cannot investigate activity in your AWS environment, responding to a security incident becomes much harder.

Minute 23–28: Amazon GuardDuty — Detect Suspicious Activity

CloudTrail helps record activity.

Amazon GuardDuty goes further by analysing supported AWS data sources to identify potentially malicious or unauthorised behaviour.

GuardDuty can detect certain indicators of:

  • Compromised AWS credentials.
  • Suspicious API activity.
  • Cryptocurrency-related abuse.
  • Unusual network communication.
  • Threats affecting supported AWS workloads.

For example, if credentials are misused from an unexpected location or exhibit suspicious API patterns, GuardDuty may generate a relevant finding.

What to configure

Open the GuardDuty console and enable the service in your AWS Region.

Review the detector configuration, available protection plans, and finding severity levels.

In a multi-account environment, centralise GuardDuty administration and coverage across relevant accounts and Regions.

GuardDuty is not a guarantee that every attack will be detected. It is one important component of a broader monitoring strategy.

Takeaway: Do not wait for customers to report suspicious behaviour before investigating your infrastructure.

Minute 28–33: AWS Security Hub — One Place to Review Security Findings

As your AWS environment grows, security findings may come from multiple services.

AWS Security Hub helps centralise supported findings and provides security posture visibility through applicable controls and standards.

Instead of investigating every service independently, security teams can review prioritised findings in a central location.

What to configure

Enable the appropriate Security Hub capabilities and review the available security standards and controls.

Pay particular attention to findings involving:

  • Excessive IAM permissions.
  • Publicly accessible resources.
  • Missing encryption controls.
  • Inadequate logging.
  • Security configuration weaknesses.

Security Hub should be combined with an agreed remediation process. A dashboard full of unresolved findings is not an effective security programme.

Takeaway: Security findings only become valuable when someone is responsible for acting on them.

Minute 33–37: AWS KMS and Secrets Manager — Protect Your Data and Credentials

Two additional services deserve attention before your first production deployment.

AWS Key Management Service (KMS) helps manage cryptographic keys used to protect data across supported AWS services.

It supports encryption strategies for services such as Amazon S3, EBS, and RDS.

AWS Secrets Manager helps securely store, retrieve, and rotate supported application secrets.

For example, instead of writing a database password directly into your application source code, you can store it in Secrets Manager and retrieve it using an authorised workload identity.

What to avoid

Never commit database passwords, AWS access keys, API tokens, or other production secrets to a public GitHub repository.

Encryption also does not replace access control. An identity with excessive permissions may still access data through authorised decryption operations.

Takeaway: Protect both the data and the credentials used to access it.

Minute 37–41: Amazon CloudWatch and AWS Budgets — Detect Problems Early

Security incidents do not always begin with obvious warnings.

Sometimes the first sign is unusual resource usage, unexpected application behaviour, or a sudden increase in spending.

Amazon CloudWatch helps monitor AWS resources, application logs, metrics, and alarms.

AWS Budgets allows you to configure cost thresholds and notifications.

For a beginner, I recommend setting up:

  • A billing budget with email notifications.
  • CloudWatch alarms for relevant application and infrastructure metrics.
  • Notifications for high-priority security findings through Amazon EventBridge and Amazon SNS.

AWS Budgets is not a real-time fraud prevention system, and billing notifications may be delayed. It should be treated as an additional warning mechanism, not a replacement for access controls or threat detection.

Takeaway: Unexpected cloud spending can be an operational warning that deserves investigation.

Minute 41–45: Amazon Inspector and AWS Config — Find Weaknesses Before Attackers Do

The final four minutes introduce two services that support continuous security assessment.

Amazon Inspector identifies supported software vulnerabilities and unintended network exposure across eligible workloads.

For example, it can help identify vulnerable packages on supported EC2 instances or vulnerabilities in container images.

AWS Config records supported resource configurations and evaluates them against defined compliance rules.

It can help detect configuration drift, such as a resource being changed from an approved security baseline.

Both services require appropriate configuration, coverage and ongoing review.

Takeaway: Security is not a one-time configuration exercise. Your infrastructure changes, and your security posture changes with it.


Your 45-Minute AWS Security Learning Roadmap

Time Service or concept What to learn
0–5 min Shared Responsibility Model Understand security ownership
5–12 min IAM MFA, roles, least privilege
12–18 min VPC and Security Groups Network isolation and firewall rules
18–23 min CloudTrail Audit API activity
23–28 min GuardDuty Detect suspicious activity
28–33 min Security Hub Review and prioritise security findings
33–37 min KMS and Secrets Manager Protect data and credentials
37–41 min CloudWatch and AWS Budgets Monitor activity and spending
41–45 min Inspector and AWS Config Assess vulnerabilities and configuration

Important: These 45 minutes are an introduction, not enough time to implement and validate every security control. Some services have usage-based costs, and multi-Region or organisation-wide configurations require additional planning.

My Minimum Security Checklist Before Going Live on AWS

Before deploying any production application, I recommend confirming the following:

  • [ ] Root user MFA is enabled.
  • [ ] Root access keys are absent.
  • [ ] Human access uses temporary credentials and least privilege.
  • [ ] Application credentials are not hardcoded.
  • [ ] Security groups expose only necessary ports.
  • [ ] Databases and sensitive resources are not unnecessarily public.
  • [ ] CloudTrail logging is configured appropriately.
  • [ ] GuardDuty monitoring covers the required accounts and Regions.
  • [ ] Security findings have owners and remediation procedures.
  • [ ] Sensitive data and secrets are protected.
  • [ ] Budget alerts and operational monitoring are configured.
  • [ ] Vulnerability scanning and configuration assessments are planned.

Final Thoughts: The Cheapest Security Investment Is Knowledge

The two client incidents I mentioned earlier reinforced an important lesson in my cloud engineering work.

You can have a talented development team, a functional application, a powerful AWS infrastructure, and a successful product launch.

But if security is overlooked during setup, the environment may contain weaknesses that remain unnoticed until an incident occurs.

The problem is not that AWS lacks security tools.

The problem is that many people start using cloud infrastructure without understanding the security responsibilities that come with it.

I am not suggesting that 45 minutes will make anyone an AWS security expert.

But 45 minutes can change the questions you ask before deployment.

It can help you recognise risky permissions, exposed resources, missing logs, and the importance of monitoring.

Before you learn how to launch your first EC2 instance, learn how to protect the AWS account that will host it.

Because in cloud computing, deploying quickly is valuable, but deploying securely is essential.


What AWS security tool do you think every beginner should learn first? Share your experience in the comments.

Tags: #AWS #CloudSecurity #Cybersecurity #DevSecOps #CloudComputing #AWSIAM #DevOps

References

  1. AWS Shared Responsibility Model
  2. AWS IAM Best Practices
  3. Amazon VPC Security
  4. AWS CloudTrail Documentation
  5. Amazon GuardDuty
  6. AWS Security Hub
  7. AWS Secrets Manager
  8. Amazon Inspector

Top comments (0)