Weekly DevOps Update: Automating the Backend Stack for Share Sampatti

Hi everyone, Here's a quick update on my DevOps progress for the Share Sampatti project at Planning Earth. This week, I focused on laying down the foundation for infrastructure automation using Terraf

What’s Done So Far

Infrastructure Setup

• Used Terraform to provision AWS EC2 instances.

• Encountered minor flakiness with Terraform responsiveness — occasional delays, and at times it doesn’t respond or hangs on apply. Currently troubleshooting these.

Application Stack with Docker Compose

• Wrote a docker-compose.yml file that successfully spins up:

  • A Node.js backend server

  • MongoDB

  • Prometheus for basic monitoring

• The command docker-compose up --build is now capable of starting the entire backend stack locally.

NGINX & Reverse Proxy

• Used Ansible playbooks to install NGINX on EC2.

• Set up a basic reverse proxy config for hosting the static app via NGINX.

What I’m Working On Next

• Automating the deployment of Docker Compose using Ansible remotely.

• Making the Ansible roles modular so they can handle:

  • Docker installation

  • Cloning project repo

  • Executing docker-compose on the remote server

• Improving Terraform stability (possibly managing backend state better or exploring remote backends like S3 + DynamoDB for state locking).

Challenges This Week

1. Slow Internet = Slow ProgressMost of my time was spent waiting for packages to install or images to build. Need to find ways to optimize or pre-pull frequently used images.

2. Terraform InstabilityInconsistent behavior during provisioning. Sometimes works, sometimes doesn’t. Still debugging if it's a provider/auth issue or network timeout.

3. Remote Docker Deployment with AnsibleInstalling Docker via Ansible is okay, but chaining that with docker-compose up on a cloned repo requires path correction and permission management.

Learnings

• IaC is powerful, but needs discipline and retries.

• Combining Terraform and Ansible gives full control — infra + config — but requires careful state and SSH management.

• Prometheus setup has helped verify that containers are up and healthy. Planning to add Grafana next.

Summary

This week’s goal of semi-automating the backend stack deployment is 50–60% achieved. The Docker Compose setup works perfectly, Ansible can configure NGINX, and Terraform can create infra — but combining all three into a single, reliable automated flow is still in progress.

Looking forward to closing the loop next week with full automation!