Deployment Guide

Complete deployment procedures for the short drama platform infrastructure.

This section covers environment setup, service deployment, configuration management, and deployment best practices for technical operations teams.

1. Overview and Architecture

This document outlines the containerized deployment procedure for the Short Drama Video-on-Demand (VOD) Platform using Docker. The system is designed with a high-availability architecture to ensure robust service delivery for streaming content.

The architecture comprises key components, including a Management Platform, a Load Balancer (HAProxy), Scalable Streaming Servers, and a highly available Main/Backup Database setup.

System Architecture Diagram

The system is structured as follows, running within a Unitel Datacenter:

1. Entry Point: Traffic enters via the Firewall.
2. Traffic Distribution: The Load Balancer (HAProxy) distributes requests to the Streaming Servers.
3. Core Services:
   • Management Platform: Backend for content, user, and data management.
   • Jump Server: Secure entry point for maintenance and internal operations.
   • Streaming Servers: Multiple instances (Nginx with RTP) for serving video streams.
   • Database: High-availability pair (Main Database and Backup Database).
   • Monitoring: Prometheus for time-series data collection, metrics storage, and alerting.

2. System Requirements

2.1 Hardware Specifications (Minimum Recommended)

These specifications are minimum recommendations. Scaling should be adjusted based on anticipated user traffic and content library size.

2.2 Software Prerequisites

3. Deployment Options

3.1 Docker Compose Deployment (Recommended for Initial Setup)

This section details the steps for a quick, localized deployment using Docker Compose. For production, consider adapting these images for Docker Swarm.

3.1.1 Initial Setup on Host Server

1. SSH into the Host Server:
   ssh user@host_ip

2. Install Docker and Docker Compose (if not present):

   • Follow the official Docker documentation for your specific Linux distribution.

3. Clone the Deployment Repository:
   git clone [https://repo.com/short-drama-vod/deployment.git](https://repo.com/short-drama-vod/deployment.git)
   cd deployment

4. Configure Environment Variables:
   Edit the .env file within the deployment directory. This file contains critical configuration details.
   # Example .env configuration
   DB_ROOT_PASSWORD=your_secure_root_password
   DB_USER=vod_user
   DB_PASSWORD=vod_secure_password
   DB_NAME=vod_platform
   HA_PROXY_PORT=80
   MANAGEMENT_API_PORT=8080
   PROMETHEUS_PORT=9090 # New port for monitoring UI

3.1.2 Executing the Deployment

Once the environment variables are set and the necessary configuration files (docker-compose.yml, config/haproxy.cfg, config/prometheus.yml) are in the current directory, use the following command to build and start all containers in detached mode (-d).

1. Build and Start Services:
   docker-compose up -d --build

   This command builds any locally defined images and starts all services (Database, Management Platform, Streaming Server, HAProxy, and Prometheus) defined in docker-compose.yml on the internal_network.

2. View Logs (Optional):
   To follow the startup progress or diagnose issues, run the logs command without the -d flag initially, or use:
   docker-compose logs -f

3.1.3 Verification

Verify that all critical containers are running and healthy.

1. Check Container Status: docker-compose ps

   Expected output should show State as Up (and ideally (healthy) for the database).

2. Test Load Balancer Access: Access the public IP or DNS of the host server on the exposed HAProxy port (default 80): curl \-I http://\<host\_ip\>

   A successful response (e.g., HTTP/1.1 200 OK) confirms HAProxy is routing traffic to the streaming servers.

3. Test Management Platform Access: The Management Platform is exposed on the host via MANAGEMENT\_API\_PORT (default 8080). curl http://\<host\_ip\>:${MANAGEMENT\_API\_PORT}/health

   Verify the API returns a successful health status code.

4. Test Prometheus Monitoring Access (New): Access the Prometheus UI on the exposed port (default 9090): curl http://\<host\_ip\>:${PROMETHEUS\_PORT}/targets

   Verify that all targets (Management Platform, Streaming Server, HAProxy) show a UP status.

3.2 Kubernetes Deployment (Production-Ready)

4. Next Steps