Deploy-It: Cloud Deployment Platform

Deploy-It is a cloud deployment platform that automates the deployment process of web applications using AWS ECS, Fargate, S3, Kafka, ClickHouse, Postgres, Docker, Node.js, and Express.

Overview

Deploy-It integrates infrastructure services and container-based workflows to streamline web app deployment, there are three main services in this project:

1. Build Server for Deployment: /build-server
2. Log Processing and Storage: /api-server
3. Request Handler Proxy: /req-handler-proxy

1. Build Server: /build-server

A containerized build executor launched on-demand via AWS ECS Fargate. This service is containerized and pushed to AWS Container Registry.

Responsibilities:

• Clone GitHub repository
• Install project dependencies
• Run production build
• Upload artifacts to S3
• Stream structured logs to Kafka (topic: build-log)

2. API Server: /api-server

An Express.js-based control plane that manages users, projects, deployments, logs, and analytics.

Core Functions:

• Project CRUD
• Trigger build deployments
• Consume and store logs from Kafka to ClickHouse
• Aggregate request analytics
• Manage user auth and session

Infra/Service Integrations:

• AWS ECS Fargate: Runs build-server container via RunTaskCommand
• Kafka: Receives structured logs from build containers
• ClickHouse: Stores structured logs for fast queries
• Prisma: ORM for PostgreSQL (users, projects, deployments)

API Routes:

POST /project

• Creates new project with a unique slug
• Input Fields: name, gitRepoURL

POST /deploy

• Accepts projectId
• Creates new deployment (status = QUEUED)
• Triggers ECS Fargate build with:
  • GIT_REPO_URL, PROJECT_ID, DEPLOYMENT_ID
  • Subnet and security group
  • Launches build-server container

GET /logs/:id

• Queries logs in ClickHouse by deployment_id

GET /analytics/:projectId

• Aggregates analytics grouped by:
  • Path, Method, User Agent, IP

POST /register & POST /login

• Registers and logs in users
• Uses bcrypt for password hashing
• Returns JWT on login

GET /projects (Auth Required)

• Lists all user-associated projects

GET /deployments/:projectId (Auth Required)

• Lists all deployments for a project

Kafka Consumer:

• Subscribes to topic build-log
• JSON-parses each log
• Inserts into log_events table in ClickHouse

3. Request Handler Proxy: /req-handler-proxy

Reverse proxy for routing frontend requests to deployed project containers.

Core Responsibilities:

• Subdomain or custom domain-based routing
• Reverse proxy to project containers
• Logs each request for analytics

Behavior:

1. Domain Parsing
   Extracts subdomain from req.hostname
   Matches with subdomain or customDomain in project table

2. Request Logging
   Asynchronously inserts:

   • path, method, userAgent, ip, projectId
     Into analytics table.

3. Proxy Handling
   Proxies request to the corresponding build files in S3.

Components

Layer	Technology
Frontend	Next.js, TailwindCSS
Backend	Node.js, Express.js, Docker
Infra	AWS ECS, Fargate, S3, ECR
Messaging	Kafka
Logs	ClickHouse
Database	PostgreSQL (via Prisma ORM)