Managing Docker Containers Like a Pro

⚙️ 1. What is a Docker Container?

A container is a running instance of a Docker image — a lightweight, isolated environment that can hold your software, tools, and files.

• Image = the blueprint (a frozen snapshot, like an OS or preconfigured setup)
• Container = the living instance of that image

When you stop or delete a container, your data and changes can vanish — unless you know how to make them persistent.

📦 2. Creating and Running a Container

To start a container interactively:

docker run -it ubuntu:22.04 bash

Here’s what each flag does:

• -i → Interactive mode (keeps STDIN open)
• -t → Allocates a pseudo-TTY (makes it feel like a real terminal)
• ubuntu:22.04 → The image to base the container on
• bash → The command to run inside (the shell)

You’ll now be inside the container’s terminal (root@containerID:/#).

🔄 3. Creating a Named Container (for Persistence)

If you don’t name your container, Docker gives it a random name. You can assign your own with --name so you can reopen it later.

docker run -it --name myubuntu ubuntu:22.04 bash

Exit when you’re done:

exit

Later, restart it and reattach with:

docker start -ai myubuntu

✅ All files and installed packages will still be there.

💾 4. Making Data Persistent (Volumes)

By default, container files live inside the container — so if it’s removed, data is gone. You can mount a directory from your host to the container using -v.

Example:

docker run -it -v ~/mydata:/root/data ubuntu:22.04 bash

• ~/mydata → folder on your host
• /root/data → folder inside the container

Now, anything you save in /root/data is stored on your host system — even if the container is deleted.

🌐 5. Mapping Ports (Accessing Apps from Host)

If your container runs a web server or service (like mdserve, Flask, etc.), you’ll need to map its ports to the host system.

Use -p to map ports:

docker run -it -p 8000:8000 ubuntu:22.04 bash

• 8000:8000 → host_port:container_port So if your app listens on port 8000 inside, you can visit it on your host at http://localhost:8000.

🧱 6. Saving Changes (Committing a Container)

When you customize a container (install software, edit configs, etc.), you can commit it to save the state as a new image.

docker commit myubuntu myubuntu:custom

Now you can start new containers from your saved configuration:

docker run -it --name myclone myubuntu:custom bash

This creates a new container with all your changes already baked in — like a snapshot.

🕹️ 7. Running a Specific Commit

Each commit you make creates a new image layer, identified by an image ID or tag.

To list your images:

docker images

Output:

REPOSITORY       TAG        IMAGE ID       CREATED          SIZE
myubuntu         custom     e21a34bfb3a4   2 minutes ago    480MB

To run a specific version by ID or tag:

docker run -it e21a34bfb3a4 bash
# or
docker run -it myubuntu:custom bash

🧹 8. Removing Containers and Images

Over time, you may accumulate old containers or images. Here’s how to clean up safely.

Remove a single container:

docker rm myubuntu

Remove all stopped containers:

docker container prune

Remove an image:

docker rmi myubuntu:custom

Remove all unused data (CAUTION):

docker system prune -a

🧩 9. Running a Container Again by Name

If you used --name, you can easily restart your container later.

Example:

docker start -ai myubuntu

That’s shorthand for:

• start → start the container
• -a → attach to it (see output)
• -i → interactive (so you can type commands)

✅ You’ll be dropped back into your previous environment instantly.

⚡ 10. Quick Reference Summary

🧠 Bonus Tips

• Use --rm only for temporary testing containers (it auto-deletes after exit).
• Always commit after large configuration changes.
• Combine volumes (-v) and ports (-p) for realistic app testing.
• Use descriptive names (--name devserver, --name docs, etc.) to stay organized.