TD1 — Run and understand one web container

Step 1 — Verify Docker and the pinned image · 20 minutes

Run these checks from your project root:

. scripts/course-env.sh
docker version
docker info --format 'server={{.Name}} os={{.OSType}}'
printf 'image=%s\nplatform=%s\n' "$NGINX_IMAGE" "$COURSE_IMAGE_PLATFORM"
docker pull --platform "$COURSE_IMAGE_PLATFORM" "$NGINX_IMAGE"
docker image inspect --format 'image_id={{.Id}}' "$NGINX_IMAGE"

Confirm that Docker reports a Linux Server and that NGINX_IMAGE contains @sha256:. Pull that reference for COURSE_IMAGE_PLATFORM, then use docker image inspect to record its local image ID. A successful pull proves image availability; it does not prove HTTP works. If the daemon is missing, start Docker Desktop or inspect the Linux Docker service before continuing.

Step 2 — Define and run one web container · 30 minutes

Create scripts/td01-run.sh yourself so a fresh Bash shell can recreate one named td01-web container from NGINX_IMAGE. Give it course.lab=TD01 and course.environment=course-lab labels, no restart policy, and a host binding of 127.0.0.1:8080 to the image's web port 8080. Use docker run --help or the Docker run reference for flag syntax. Check your script with bash -n, execute it, then request http://127.0.0.1:8080/ using curl and your browser. Save the real response under evidence/TD01/. If the name or port is occupied, inspect its owner and labels before changing anything.

Step 3 — Inspect the process and network path · 30 minutes

Use these ordinary inspection commands, then save the outputs that support your claim:

docker ps --filter name=td01-web
docker ps -a --filter name=td01-web
docker inspect --format 'id={{.Id}} state={{.State.Status}} image={{.Config.Image}}' td01-web
docker top td01-web
docker port td01-web
docker logs --tail 30 td01-web

Save the actual image reference, container ID, state, processes, port bindings and one HTTP result. Explain in your own words which command proves that the process runs, which proves the host binding, and which proves an HTTP request succeeds. Check that the host IP is loopback; a wildcard binding does not satisfy this lab.

Step 4 — Predict, stop, start, remove and recreate · 55 minutes

Before each change, write a one-line prediction for the container ID, docker ps -a state, HTTP result and image availability. Save the original ID. Stop the container and check its state and HTTP failure. Start the same object and compare its ID with the original. Then stop and remove only your labelled TD1 object, confirm the image remains, and run your saved script to create a new object. Compare the new ID and HTTP response. Save actual outputs, including a failed request; do not write expected values into evidence files.

Step 5 — Check and leave a clean host · 45 minutes

Explain why start preserved the ID while recreation changed it. Confirm that the recreated container uses the pinned image and serves HTTP. Stop and remove only td01-web, leaving the image and your saved script in place for review. Commit the script to your project repository and write evidence/TD01/OBSERVATIONS.md with the commands, observed IDs, HTTP results, and any failed check. Do not use docker system prune or remove unrelated objects. TD2 will use port 8080 for your API image.