Automated Dev Prod Deployments

Automated Deployment Steps

This page outlines the procedure for deploying code automatically using environment-specific deployment scripts. These scripts replace the manual deployment process (temp process) and ensure consistent, reliable, and faster deployments to all UnlockEd environments.

---

Table of Contents

• Purpose
• Overview
• Deployment Scripts Location
• Supported Environments
• How It Works
• Quick Start
• Command Line Options
• Validation
• Troubleshooting

---

Purpose

• Automate Deployments: Streamline the end-to-end deployment process for all environments.
• Ensure Consistency: Enforce a standardized deployment pattern across Demo and Production systems.
• Increase Speed: Deploy code in minutes rather than hours.

---

Overview

Each environment (Demo, Alaska, Maine, St. Louis, MoCode) has its own script that handles:

1. Git Operations – Cloning, branch checkout, merge, and push
2. Docker Build & Push – Building backend, frontend, and middleware images to AWS ECR
3. Kubernetes Deployment – Restarting deployments and monitoring rollout
4. Port Forwarding – Automatically establishing cluster access for production environments

---

Deployment Scripts Location

All deployment scripts are located in the UnlockedLabs shell_scripts repository:

📁 GitHub Repository:
https://github.com/UnlockedLabs/shell_scripts/tree/main/deploment_scripts

Each script corresponds to a specific environment:

Script Name	Environment	Branch	Tag	Type
deploy-demo.sh	Demo (Development)	demov2	demo	Development
deploy-alaska.sh	Alaska	alaska	alaska	Production
deploy-maine.sh	Maine	maine	maine	Production
deploy-stlouis.sh	St. Louis	stlouis	stlouis	Production
deploy-mocode.sh	MoCode	mocode	mocode	Production

---

Supported Environments

Environment	Branch	Namespace	Context
Demo (Development)	demov2	demo	stgdmo
Alaska	alaska	default	alaska
Maine	maine	default	maine
St. Louis	stlouis	default	stlouis
MoCode	mocode	default	mocode

---

How It Works

Each script executes the same general pipeline:

1. Git Operations

• Clone the repository
• Check out the environment-specific branch
• Fetch latest changes and hard reset to origin/main
• Force push updates to the environment branch

2. Docker Image Build & Push

Build and push three container images to AWS ECR using docker buildx:

• Backend → unlockedv2:<tag>
• Frontend → frontend:<tag>
• Provider Middleware → provider_middleware:<tag>

3. Kubernetes Rollout

• Switch to the correct kubectl context
• Restart server, frontend, and provider-service deployments
• Wait for rollout completion (up to 10,000 seconds)

4. Port Forwarding (Production Only)

• Sets up automated port forwarding via ~/.local/bin/<env>_port_forward.sh
• Logs available at <env>_port_forward.log

---

Quick Start

Deploy to Demo

./deploy-demo.sh

Deploy to a Production Environment

./deploy-alaska.sh
./deploy-maine.sh
./deploy-stlouis.sh
./deploy-mocode.sh

---

Command Line Options

Option	Description
--workdir <path>	Custom working directory (default: /tmp)
--skip-git	Skip Git operations
--skip-build	Skip Docker build & push
--skip-deploy	Skip Kubernetes rollout
--dry-run	Show planned actions without executing
-h, --help	Display help message

Example:

./deploy-alaska.sh --skip-build --dry-run

---

Validation

After a successful deployment:

1. Check Kubernetes Status:

   kubectl get pods
   kubectl get deployments

2. Verify Application Health:

   • Test production endpoints
   • Confirm new image tags are running

3. Monitor Logs:

   kubectl logs -f deployment/server
   kubectl logs -f deployment/frontend
   kubectl logs -f deployment/provider-service

---

Troubleshooting

Common Issues

Issue	Resolution
Missing AWS Credentials	Run aws configure
Missing Docker Buildx	Run docker buildx install
Kubernetes Context Missing	Run kubectl config get-contexts and verify
Port Forwarding Conflicts	Kill process with `lsof -t -i :6443

Clean Up Temporary Files

rm -rf /tmp/*-deploy-*