Container Lifecycle — Run, Stop, Start, Remove Explained

What You'll Learn

Beginner

• The five states every container goes through
• Which commands transition between states
• The difference between stop, kill, and pause
• How to clean up containers properly

The Five Container States

State 1: created

The created state means a container exists but hasn't started. This is rarely used directly — docker run combines create + start. But you can do them separately:

# Create a container without starting it
docker create --name my-container ubuntu

# It exists but isn't running
docker ps -a
# CONTAINER ID   IMAGE    COMMAND   CREATED         STATUS    NAMES
# abc123         ubuntu   "bash"    3 seconds ago   Created   my-container

# Start it later
docker start my-container

State 2: running

The running state is when the container's main process (PID 1) is executing. This is the normal state for active containers.

# Start a container in running state
docker run -d --name web nginx

# Check it's running
docker ps
# STATUS: Up 5 seconds

How a container stays running

A container runs as long as its main process (PID 1) is alive. If the process exits, the container stops. This is why:

• docker run ubuntu echo hello — runs echo, prints "hello", exits immediately.
• docker run -d nginx — nginx's master process runs forever, container stays up.
• docker run -it ubuntu bash — bash runs as long as you're in the shell. Exit → container stops.

State 3: paused

The paused state freezes the container's processes in memory. The container is still using RAM but isn't executing. Useful for temporarily freeing CPU without stopping the container.

# Pause a running container
docker pause web

# Check status
docker ps
# STATUS: Up 2 minutes (Paused)

# Unpause to resume
docker unpause web

State 4: stopped (exited)

A container enters the stopped state when:

• The main process exits naturally.
• You run docker stop (graceful shutdown).
• You run docker kill (immediate kill).

docker stop vs docker kill

# Graceful stop (recommended)
docker stop web

# Force kill (if stop hangs)
docker kill web

# Stop with custom timeout
docker stop -t 30 web   # wait 30 seconds before SIGKILL

A stopped container still exists!

# See stopped containers
docker ps -a
# STATUS: Exited (0) 2 minutes ago

# Restart a stopped container
docker start web

# Remove a stopped container
docker rm web

State 5: removed

The removed state is final — the container and its filesystem are deleted. This cannot be undone.

# Remove a stopped container
docker rm web

# Force remove a running container (stops then removes)
docker rm -f web

# Remove all stopped containers
docker container prune

The Full Lifecycle in Practice

Auto-Remove with --rm

The --rm flag tells Docker to remove the container automatically when it stops. Perfect for one-off commands:

# Container is removed as soon as the command finishes
docker run --rm alpine echo "hello"

# Great for one-off tasks
docker run --rm -v $(pwd):/data alpine ls /data

# NOT suitable for long-running services you want to restart later

Restart Policies

For production containers, you want them to restart automatically if they crash or if the host reboots. Use --restart:

# Always restart (even on reboot)
docker run -d --restart always nginx

# Restart unless explicitly stopped
docker run -d --restart unless-stopped nginx

# Restart up to 3 times on failure
docker run -d --restart on-failure:3 nginx

Common Mistakes

Practical Exercise (10 minutes)

Mini Challenge

Run a container with --restart unless-stopped. Stop it explicitly with docker stop. Restart the Docker daemon (or restart your computer). Does the container come back? (Hint: with unless-stopped, it should NOT come back because you explicitly stopped it. Try the same with --restart always to see the difference.)