Week 5-6: Cloud Automation, Cost Optimization, and Serverless Ops with AWS Lambda & Boto3

What actually went down this week

Weeks 5 and 6 focused heavily on cloud foundations, automation, monitoring, and cost optimization — connecting learned concepts into practical scripts and serverless automation.

I built multiple real-world AWS automation projects including:

• Automating AWS resource provisioning with Python Boto3: EC2 launch, Apache setup, S3 bucket/file management, IAM policy attachments, and CloudWatch alarms.

• Hands-on AWS Lambda function implementing cleanup of stale EBS snapshots to reduce storage costs. Scheduled with EventBridge cron.

• Setup IAM least privilege roles for Lambda and EC2 automation scripts for security best practices.

• Detailed log monitoring and troubleshooting using CloudWatch Logs for Lambda executions and EC2 monitoring alarms.

---

Why I had to tackle these topics now

Moving beyond manual AWS Console interaction towards scalable automation is critical for real DevOps/Cloud proficiency:

• Automating provisioning makes infrastructure repeatable and error-resistant.

• Lambda-based cleanup helps control AWS costs proactively.

• IAM best practices reduce security risks in automation workflows.

• CloudWatch monitoring enables real-time and historical insight on resource health.

---

The topics I finished (short notes + useful commands)

AWS Automation with Boto3

Scripts to launch EC2, install Apache, create S3 buckets, and setup CloudWatch alarms programmatically:

python scripts/automation.py

Key AWS CLI commands monitored/logged within scripts.

AWS Lambda Snapshot Cleaner

Python code automates deletion of snapshots orphaned from volumes or instances:

# Fetch snapshots and instances, delete stale snapshots 

def lambda_handler(event, context):

Scheduled using EventBridge cron for weekly run.

IAM Role & Policy Management

Configured AWS IAM roles with well-scoped policies for Lambda and EC2 automation.

CloudWatch Monitoring & Logs

Used CloudWatch Logs to monitor Lambda runs and EC2 CPU alarms.

---

The things that broke (and how I fixed them)

• Lambda permission errors → Enhanced IAM execution role

• Snapshot volume missing edge cases → Added exception handling

• Timeout in Lambda for large snapshot sets → Increased timeout setting

• Cron environment quirks → Added absolute paths and logging in scripts

---

What I learned the hard way

• Infrastructure as code is key to reliable cloud ops.

• Lambda automation can reduce AWS costs dramatically.

• Least-privilege IAM roles critical for security.

• Logs and alarms are vital troubleshooting tools.

• Scheduling with EventBridge ensures continuous automation.

---

Resources I actually used

• AWS Lambda Docs

• Boto3 Documentation

• AWS IAM Best Practices

• CloudWatch Scheduler & Logs

---

Up Next: Week 7 Plan

• Deep dive into AWS networking: VPCs, subnets, route tables, and gateways

• Build public-private subnet architectures with bastion hosts for secure access - Create layered security using Security Groups and Network ACLs (NACLs)

• Explore Terraform basics: install, configure AWS CLI, deploy networking and EC2

• Automate IAM user/role management with Boto3 and configure AWS Budgets for cost control

• Launch scalable EC2 instances using Terraform and EC2 Launch Templates with IAM roles attached

---

Follow along

GitHub: shauryad01/cloud-devops-journey
LinkedIn: Shaurya Dhingra