Example: from servers to containers

Example with made-up data

$ git log --oneline container-migration

From servers to containers: what changes

Rossi Meccanica’s business app moves from a hand-configured server to containers described in a file: every change is tested, approved and released in minutes, and rolling back is one command.

Architecture before and after

BEFORE

AFTER

A real CI/CD pipeline (example)

pipeline #214 · “fix: VAT on credit notes” · branch main

code checkout3 s
automated tests (182 tests)1 min 12 s
image build48 s
vulnerability scan: 0 critical22 s
deploy to test environment15 s
approval: Marco R. (manager)–
deploy to production, no downtime18 s
health check: answers 200 OK5 s

Total: 3 min 3 s of automation + 1 approval click.

A commented docker-compose excerpt

# docker-compose.yml – Rossi Meccanica business app (excerpt)
services:
  proxy:
    # automatic HTTPS, certificates renew themselves
    image: caddy:2.8
    # the only port open to the outside
    ports: ["443:443"]
  app:
    # exact version: rolling back = changing this number
    image: registry.rossi-meccanica.example/gestionale:1.14.2
    # passwords kept out of the code, never in the repository
    env_file: .env
    depends_on:
      db: { condition: service_healthy }
    # if it hangs, it restarts by itself
    restart: unless-stopped
  db:
    image: postgres:16
    # data survives even when the container is recreated
    volumes: ["db-data:/var/lib/postgresql/data"]
    healthcheck:
      test: ["CMD", "pg_isready", "-U", "gestionale"]
  backup:
    image: registry.rossi-meccanica.example/pg-backup:1.3
    # database dump every night at 1:00, then off site
    environment: { SCHEDULE: "0 1 * * *" }

Esempio con dati fittizi

$ git log --oneline migrazione-container

Da server a container: come cambia

Il gestionale di Rossi Meccanica passa da un server configurato a mano a container descritti in un file: ogni modifica viene provata, approvata e pubblicata in pochi minuti, e si torna indietro con un comando.

Architettura prima e dopo

PRIMA

DOPO

Una pipeline CI/CD vera (esempio)

pipeline #214 · “fix: calcolo IVA sulle note di credito” · ramo main

checkout del codice3 s
test automatici (182 test)1 min 12 s
build dell’immagine48 s
scansione vulnerabilità: 0 critiche22 s
deploy in ambiente di test15 s
approvazione: Marco R. (responsabile)–
deploy in produzione, senza fermo18 s
controllo di salute: risponde 200 OK5 s

Totale: 3 min 3 s di automazione + 1 clic di approvazione.

Un estratto del docker-compose, commentato

# docker-compose.yml – gestionale Rossi Meccanica (estratto)
services:
  proxy:
    # HTTPS automatico, certificati rinnovati da solo
    image: caddy:2.8
    # l’unica porta aperta verso l’esterno
    ports: ["443:443"]
  app:
    # versione precisa: tornare indietro = cambiare questo numero
    image: registry.rossi-meccanica.example/gestionale:1.14.2
    # password fuori dal codice, mai nel repository
    env_file: .env
    depends_on:
      db: { condition: service_healthy }
    # se si blocca, riparte da solo
    restart: unless-stopped
  db:
    image: postgres:16
    # i dati restano anche se il container viene ricreato
    volumes: ["db-data:/var/lib/postgresql/data"]
    healthcheck:
      test: ["CMD", "pg_isready", "-U", "gestionale"]
  backup:
    image: registry.rossi-meccanica.example/pg-backup:1.3
    # copia del database ogni notte all’1:00, poi fuori sede
    environment: { SCHEDULE: "0 1 * * *" }

← Back to services ← Torna ai servizi