K3s Application Platform

Self-managed infrastructure used to implement and explore production-style deployment patterns for containerized Next.js applications, from local builds through Kubernetes workloads and private reverse-proxy access.

DockerKubernetesK3sPostgreSQLCaddyUbuntu

Abstract container deployment flow through K3s, a Kubernetes service, private networking, and Caddy

Context

The Problem

Containerized Next.js workloads needed a repeatable path from development builds to a self-managed Kubernetes environment, including database connectivity, private ingress, deployment automation, and operational troubleshooting.

System flow

Architecture

Developer / Git

Docker Build

K3s Deployment

Kubernetes Service

Tailscale Network

Caddy

Application

Implementation

Engineering

Dockerized Next.js applications and used Docker Compose for the development workflow.
Configured K3s workloads with Kubernetes Deployments and Services, including local container image deployment and rolling deployments.
Built Bash-based deployment automation and managed application environment variables.
Connected Kubernetes pods to PostgreSQL and exposed applications through Caddy over a Headscale/Tailscale private network.
Debugged containers and pod lifecycles while operating the underlying Ubuntu host.

What required care

Challenges

Diagnosing Kubernetes DiskPressure and expanding Linux storage with LVM.
Troubleshooting pod lifecycle and local container image behavior.
Maintaining PostgreSQL connectivity and environment configuration from inside Kubernetes pods.

Technology

Stack

Status: Personal Project

DockerKubernetesK3sPostgreSQLCaddyUbuntu