Praxis

This project demonstrates a complete DevOps lifecycle for deploying OpenWebUI on Microsoft Azure — starting from infrastructure provisioning using Terraform, followed by configuration management, containerization, and fully automated CI/CD using Jenkins.

The setup reflects real-world production practices, not a demo or tutorial-only workflow.

---

🔁 High-Level Flow

Terraform → Azure VM
        ↓
Ansible → VM Configuration
        ↓
Docker → Application Container
        ↓
Jenkins → CI/CD Automation
        ↓
Docker Hub → Image Registry
        ↓
Azure VM → Production Deployment

---

🧱 Phase 1 – Infrastructure Provisioning (Terraform)

Objective

Provision all Azure infrastructure from scratch using Terraform.

Tools

• Terraform
• Azure CLI
• Azure Resource Manager (ARM)

Resources Created

• Resource Group
• Virtual Network
• Subnet
• Network Security Group (NSG)
• Public IP
• Network Interface
• Linux Virtual Machine (Ubuntu 22.04)

Key Concepts

• Infrastructure as Code (IaC)
• Idempotent deployments
• Declarative provisioning

Steps

1. Authenticate with Azure:

   az login

2. Create a Service Principal for Terraform:

   az ad sp create-for-rbac --role Contributor --scopes /subscriptions/<SUB_ID>

3. Export credentials:

   export ARM_CLIENT_ID=...
   export ARM_CLIENT_SECRET=...
   export ARM_TENANT_ID=...
   export ARM_SUBSCRIPTION_ID=...

4. Initialize and apply Terraform:

   terraform init
   terraform plan
   terraform apply

Result: ✅ Azure VM with public IP and secured networking.

---

⚙️ Phase 2 – VM Access & Base Setup

Objective

Prepare the VM for automation and container workloads.

Steps

• SSH into VM using key-based authentication

• Verify ports:

  • 22 → SSH
  • 80 → HTTP
  • 443 → HTTPS
  • 3000 → OpenWebUI
  • 8081 → Jenkins

Result: ✅ VM ready for configuration management.

---

🛠️ Phase 3 – Configuration Management (Ansible)

Objective

Automate VM configuration and application setup.

What Ansible Handles

• Package installation
• Docker installation
• Docker service management
• Environment file setup
• Persistent volume setup
• Container execution

Why Ansible

• No agent required
• Repeatable and predictable
• Ideal for single-node servers

Result: ✅ OpenWebUI runs in Docker with persistent storage.

---

📦 Phase 4 – Containerization (Docker)

Objective

Run OpenWebUI as a containerized service.

Key Details

• Docker image: OpenWebUI (custom build later via CI)

• Persistent data mounted at:

  /opt/openwebui-data
  

• Environment variables stored securely on the VM

Result: ✅ Application survives restarts and redeployments.

---

🌐 Phase 5 – Domain, Nginx & SSL

Objective

Expose OpenWebUI securely over HTTPS.

Steps

• Domain mapped via DNS (A record → VM public IP)
• Nginx configured as reverse proxy
• SSL enabled (HTTPS)

Result: ✅ OpenWebUI accessible via custom domain over HTTPS.

---

👥 Phase 6 – Multi-User Control (Application Level)

Objective

Enable and manage multiple users securely.

Features

• Admin and user separation
• Default model control
• System-level prompts enforced
• Users cannot override admin policies

Result: ✅ Controlled multi-user environment.

---

🔁 Phase 7 – CI/CD Automation (Jenkins)

Objective

Achieve fully automated CI/CD on every production push.

---

🔹 CI (Continuous Integration)

Triggered on push to master branch.

Pipeline tasks:

1. Clone GitHub repository
2. Build Docker image
3. Tag image (latest)
4. Push image to Docker Hub
5. Cleanup build cache

---

🔹 CD (Continuous Deployment)

After CI succeeds:

1. Jenkins SSHs into Azure VM
2. Executes deployment script
3. Pulls latest Docker image
4. Replaces running container
5. Rolls back automatically if failure occurs

Deployment script location:

/opt/openwebui/deploy.sh

---

🔐 Security Measures

• SSH key-based authentication only
• known_hosts enforced (no insecure SSH flags)
• Secrets stored outside GitHub
• Jenkins credentials stored securely
• No hardcoded tokens in pipeline

---

♻️ Rollback Strategy

If deployment fails:

• Old container is preserved
• New container is removed
• Previous version is restored automatically

Result: ✅ No downtime, no manual intervention.

---

🌿 Branch Strategy

Branch	Purpose	CI/CD
dev	Development	❌
master	Production	✅

Only master triggers deployment.

---

📌 What This Project Proves

• Terraform-based Azure provisioning
• Real CI/CD (not manual builds)
• Secure Jenkins automation
• Docker image lifecycle management
• Production rollback handling
• End-to-end DevOps ownership

---

🚀 Future Enhancements

• Blue-Green deployment
• Monitoring & alerts
• Secrets via Azure Key Vault
• AKS migration

---

👤 Author

Dharshan Kumar Built as a hands-on, production-focused DevOps implementation.

---