chatapp-infrastructure

IaC repository for deploying SlurmTalks (a Twitter-like social platform) to a Kubernetes cluster. Covers application containerisation, Kubernetes orchestration via Helm, and Prometheus metrics scraping.

Tech Stack

IaC & Deployment Tools

Tool	Role
Docker	Application containerisation (multi-stage Alpine builds)
Helm	Kubernetes package management and release templating
Kubernetes	Container orchestration

Application Components

Component	Technology	Port
Frontend	Next.js 15 / React 19 / TypeScript / Tailwind CSS (Node 20)	3000
Backend	Go 1.24 / Gin	8080
Database	PostgreSQL	5432
Cache	Redis	6379

Observability

Component	Technology
Metrics	Prometheus (scraped from backend /metrics)

Prerequisites

• Docker
• kubectl configured for the target cluster
• Helm ≥ 3.x
• Running PostgreSQL and Redis instances accessible from the cluster

Configuration

# Override Helm values for your environment
cp helm/slurmtalks/values.yaml helm/slurmtalks/values.local.yaml
# Edit: image tags, ingress hostname, DB/Redis connection details

Key Variables

Variable	Source	Description	Example
POSTGRES_HOST	ConfigMap	PostgreSQL host	postgres.slurmtalks.svc.cluster.local
POSTGRES_PORT	ConfigMap	PostgreSQL port	5432
POSTGRES_DB	Secret	Database name	slurmtalks
POSTGRES_USER	Secret	Database user	app
POSTGRES_PASSWORD	Secret	Database password	s3cr3t
REDIS_ADDR	ConfigMap	Redis address with port	redis-master:6379
REDIS_PASSWORD	Secret	Redis password	r3dis
JWT_SECRET_KEY	Secret	JWT signing secret	changeme
NEXT_PUBLIC_BACKEND_URL	ConfigMap	Backend API URL seen by frontend	https://host/api
MIGRATE_IN_CODE	ConfigMap	Run DB migrations on startup	true
SIMULATED_METRICS	ConfigMap	Enable fake metric generation	false

Project Structure

chatapp-infrastructure/
├── backend-service/   # Go API — auth, tweets, users; Dockerfile; SQL migrations
├── frontend-service/  # Next.js UI; Dockerfile
├── helm/              # Helm umbrella chart (slurmtalks) with backend & frontend sub-charts
├── longhorn/          # Longhorn persistent storage configuration
└── README.md

Architecture

graph TD
    Dev["👨‍💻 Developer"]
    Dev -->|"docker build + push"| Registry["🗂️ Container Registry"]
    Dev -->|"helm upgrade"| Cluster

    subgraph Cluster["☸️ Kubernetes Cluster"]
        Ingress["🔀 NGINX Ingress\nslurm-talks-stage.s023943.edu.slurm.io"]

        subgraph App["📦 slurmtalks namespace"]
            FE["🌐 Frontend\nNext.js :3000"]
            BE["⚙️ Backend\nGo/Gin :8080"]
            PG[("🐘 PostgreSQL\n:5432")]
            RD[("⚡ Redis\n:6379")]
        end

        subgraph Observability["📊 Observability"]
            Prom["🔭 Prometheus"]
        end
    end

    Registry -->|"image pull"| App
    Ingress -->|"/"| FE
    Ingress -->|"/api"| BE
    FE --> BE
    BE --> PG
    BE --> RD
    BE -.->|"/metrics scrape"| Prom

Loading

Ingress Routing

Path	Service	Port
/	frontend	80
/api	backend	8080

Default hostname: slurm-talks-stage.s023943.edu.slurm.io

Resource Limits

Component	CPU Request	CPU Limit	Memory Request	Memory Limit	HPA
Backend	250m	500m	256Mi	512Mi	1–3 replicas (CPU 70%)
Frontend	100m	250m	128Mi	256Mi	—