Docker Configuration Evolution: From Scattered Containers to Managed Stacks

Mon, 01 Jan 0001 00:00:00 +0000

Previously, each service lived in a separate repository with minimal docker-compose.yml:

# Old approach: garrysmod-server/
services:
 garrysmod-server:
 image: ceifa/garrysmod:latest
 ports:
 - "27015:27015"
 volumes:
 - ./garrysmod:/data
 restart: unless-stopped

Problems:

❌ No resource control - one service could consume all memory
❌ No health checks - crashed services went undetected
❌ No log rotation - logs grew until disk full
❌ No network isolation - all services in default network
❌ Manual per-service updates - high risk of human error

New Approach: Stacks with Explicit Contracts
#

Now related services are grouped into a single repository with unified configuration:

Many Projects. One Structure. No Chaos.

Mon, 01 Jan 0001 00:00:00 +0000

When you have more than ten projects, questions arise:

Where is the config for this service?
Which environment variable to set on deploy?
Why does one container consume all memory while another stays silent on crash?
How to quickly deploy all this on a new machine?

Answer: unified structure, unified rules, unified approach.

🗂 Unified Repository Structure
#

Every active project follows one template:

Infrastructure on Ponfertato

Docker Configuration Evolution: From Scattered Containers to Managed Stacks

New Approach: Stacks with Explicit Contracts #

Many Projects. One Structure. No Chaos.

🗂 Unified Repository Structure #

New Approach: Stacks with Explicit Contracts
#

🗂 Unified Repository Structure
#