AWS Python Automation Suite

Production-grade AWS operations automation using Python and Boto3 — EC2 control, S3 management, Lambda scheduling, CloudWatch monitoring, and SNS alerting.

---

What This Project Does

A Python Boto3 automation suite that replaces manual AWS console operations with code-driven workflows. Covers the five most common cloud operations tasks: EC2 instance lifecycle management, S3 file operations, Lambda function scheduling, CloudWatch alarm creation, and SNS email alerting. Each module is independently usable and production-ready.

---

Architecture

         ┌─────────────────────────────────────────┐
         │           main.py (Orchestrator)         │
         └──┬──────────┬──────────┬────────────┬────┘
            │          │          │            │
   ┌────────▼──┐ ┌─────▼──────┐ ┌▼─────────┐ ┌▼──────────────┐
   │ec2_       │ │s3_         │ │lambda_   │ │cloudwatch_    │
   │controller │ │uploader    │ │scheduler │ │alerts         │
   │           │ │            │ │          │ │               │
   │Start/Stop │ │Upload/List │ │Schedule  │ │Create Alarms  │
   │EC2        │ │S3 Objects  │ │Lambda    │ │CPU/Cost/Error │
   │Instances  │ │            │ │Invocations│ │+ SNS Email   │
   └─────┬─────┘ └─────┬──────┘ └────┬─────┘ └───────┬───────┘
         │             │             │               │
         └─────────────┴─────────────┴───────────────┘
                                 │
                        AWS Cloud (Boto3 SDK)

---

Modules

Module	AWS Service	What It Does
ec2_controller.py	EC2	Start, stop, list instances by tag or region
s3_uploader.py	S3	Upload files, list bucket contents, generate presigned URLs
lambda_scheduler.py	Lambda + EventBridge	Schedule Lambda functions, manage invocation rules
cloudwatch_alerts.py	CloudWatch + SNS	Create CPU, cost, and error rate alarms with email alerts
main.py	Orchestrator	Runs all modules in sequence with configurable parameters

---

Nokia OAM → AWS Operations Mapping

This toolkit applies Nokia OAM (Operations, Administration, Maintenance) principles directly to cloud operations — the same operational discipline used to manage 5G network infrastructure, applied to AWS resource management.

Nokia OAM Function	AWS Automation Equivalent
Node health monitoring	CloudWatch alarms + SNS alerts
Resource provisioning	EC2 controller (start/stop/list)
Data collection	S3 uploader (logs, reports, configs)
Event scheduling	Lambda scheduler + EventBridge
Fault management	CloudWatch error rate alarms

---

Security Design

• IAM least privilege — Boto3 uses AWS credentials with minimum required permissions
• No hardcoded credentials — all credentials loaded from environment variables or AWS CLI config
• SNS topic scoping — alerts sent only to verified email endpoints
• S3 presigned URLs — time-limited access to objects, no permanent public exposure

---

Key Design Decisions

Why five separate modules instead of one script? Each module has a single responsibility — the same principle Nokia uses for network function separation. ec2_controller.py only manages EC2. If EC2 logic changes, only one file changes. This is maintainable, testable, production-grade code.

Why CloudWatch alarms instead of manual checking? Manual console monitoring does not scale. CloudWatch alarms with SNS notifications mean issues are detected and reported automatically — the same proactive fault detection Nokia OAM applies to 5G network nodes.

Why Lambda scheduling over cron jobs? Lambda + EventBridge is serverless, managed, and requires zero server maintenance. Cost is near-zero for scheduled tasks. No EC2 instance running 24/7 just to run a cron job.

---

Quick Start

# Clone
git clone https://github.com/sadvi11/aws-python-automation.git
cd aws-python-automation

# Install dependencies
pip install -r requirements.txt

# Configure AWS credentials
aws configure

# Run all modules
python main.py

# Or run individual modules
python ec2_controller.py
python s3_uploader.py
python cloudwatch_alerts.py
python lambda_scheduler.py

---

Example Output

EC2 Controller:
  ✓ Instance i-0abc123 → running
  ✓ Instance i-0def456 → stopped

S3 Uploader:
  ✓ Uploaded: report.csv → s3://my-bucket/reports/
  ✓ Generated presigned URL (valid 1 hour)

CloudWatch Alerts:
  ✓ CPU alarm created: threshold 80%
  ✓ SNS notification: sadhvisharma763@gmail.com

Lambda Scheduler:
  ✓ Rule created: daily-cleanup → 0 2 * * ? *

---

Repository Structure

aws-python-automation/
├── main.py                 # Orchestrator — runs all modules
├── ec2_controller.py       # EC2 start/stop/list automation
├── s3_uploader.py          # S3 file operations and presigned URLs
├── lambda_scheduler.py     # Lambda scheduling with EventBridge
├── cloudwatch_alerts.py    # CloudWatch alarms + SNS email alerts
├── requirements.txt        # boto3, python-dotenv
├── .gitignore
└── README.md

---

Design Decisions

• Boto3 vs AWS CLI — when to use each, why Python automation beats manual console work
• IAM least privilege — what permissions each module needs and why
• CloudWatch alarms — threshold types, alarm states, SNS integration
• EventBridge vs CloudWatch Events — the evolution and when each applies
• S3 presigned URLs — how they work, expiry, use cases
• Lambda invocation types — synchronous vs asynchronous vs event-driven

---

Author

Sadhvi Sharma — Cloud & AI Engineer Nokia India (5G Packet Core) → Cloud & AI Engineering Calgary, AB, Canada | Permanent Resident | Open to Relocation

LinkedIn | GitHub

Canadian Compliance Note

All resources deploy to ca-central-1 by default. Supports PIPEDA data residency requirements for Canadian workloads.