Lab 20: Bad image deployment and rollback

Objective

Walk through the canonical “new deploy is broken, roll back” path.

Tasks

Task 1: Bring up a known-good version

mkdir -p ~/rollback-lab && cd ~/rollback-lab

cat > compose.yml <<'EOF'
services:
  web:
    image: myorg/web:1.0.0
    ports: ["8080:80"]
    healthcheck:
      test: ["CMD", "wget", "-qO-", "http://localhost"]
      interval: 5s
      retries: 3
EOF

# Use a real image as the "good" baseline
sed -i 's/myorg\/web:1.0.0/nginx:1.27/' compose.yml

docker compose up -d
sleep 5
curl -s http://localhost:8080 | head -1
# <!DOCTYPE html>...
docker compose ps

Task 2: Deploy a broken image

sed -i 's/nginx:1.27/myorg/web:does-not-exist/' compose.yml

docker compose up -d
sleep 5
docker compose ps
# web       myorg/web:does-not-exist   ... Exit 1
docker compose logs web
# Error response from daemon: pull access denied ... or no such image

Task 3: Detect via health

The container is in Exit 1. A monitoring system would alert here. For the lab, observe the state directly:

docker inspect $(docker compose ps -q web) --format '{{.State.Status}} {{.State.ExitCode}}'

Task 4: Roll back

sed -i 's/myorg\/web:does-not-exist/nginx:1.27/' compose.yml
docker compose up -d
sleep 5
docker compose ps
curl -s http://localhost:8080 | head -1

Task 5: Verify the previous image is the running one

docker compose ps --format json | jq -r '.[].Image'
# nginx:1.27

Cleanup

docker compose down

Verification status

Last reviewed

2026-08-09

Executed end to end

not yet run on hardware

The commands and configuration here have been reviewed against the verified software versions, but nobody has run this lab start to finish on a system meeting its prerequisites. Treat the Expected Outcome as the intended result rather than an observed one, and keep the Cleanup section to hand.

Objectives

Prerequisites