Docker Tutorial - Step by Step Guide from the Beginner to Advance

Published on Oct. 16, 2021

In order to scale up the production, industries are adopting DevOps practices. DevOps is not only providing rapid application development but also a high level of security.  This can be ensured by the involvement of container technology in DevOps tools. And one of the very famous container tools is Docker. 

 

In this article, we are going to learn  Docker step by step from the very beginning to the advanced level. You will find almost all the commands and their concepts that we generally used while working on Docker. After going through this article, you would be able to find Docker Tutorial:

• Containerize almost any application
• Upload custom Docker images to Docker Hub(online registry)

 

Table of Content

• What is docker?
• Getting started with Docker
  • Installation of Docker
• Basic terminology of Docker
  • Docker image
  • Docker Hub
  • Dockerfile
  • Docker run
  • Docker Engine
• How to run a container?
• How to manage your Docker container?
  • Container closed while exiting
  • How to stop a container?
  • How to restart your container?
  • How to again go inside the container?
  • How can you execute a command without going inside your container?
  • How to check the logs of your container
• Docker Networking
  • How to create our own network?
  • LoadBalancing using DNS
  • Network Address Translation(NAT) in Docker
• Docker Volume
  • Benefits of Persistent storage
  • How to add Persistent storage in Docker container
• How to Pass variable to our Docker Container
• How to create own Custom Docker Image
  • What is Docker Image?
  • Why create your own Docker image
  • Docker commit
  • Dockerfile
• Final Voice

 

Docker is an open platform for developing, shipping, and running applications. Docker enables you to separate your applications from your infrastructure so you can deliver software quickly. With Docker, you can manage your infrastructure in the same ways you manage your applications. To know more about Docker follow this link.

 

In order to start the tutorial on Docker, you will need Docker software to be installed in your system. If you are using AWS cloud and inside your EC2 you may not need to install the Docker there.

 

In this article, We are not going to discuss the installation of Docker. You can find the complete guide on the official website for the installation-

Installation of Docker

 

To continue our learning there is some basic terminology of Docker that we should know. Let's start one by one.

 

To run the container just in one second images are the one who is responsible. Docker images are like an entire os. It is created with minimal functionality of an application or OS. There are some pre-created images we can use or we can create our own as well.

 

Docker Hub is an online community of Docker where you can find all the available images. You can also upload your own image to this and contribute to the world. If you are the one who works for opensource it is a great place for you. It is super easy to upload and update your images there.

 

A Dockerfile is a text file having all the details about the configuration of the image. We can create our own image using the Dockerfile. There might be several occasions we required manual setup of environment there we can use the Dockerfile to create a custom image. Once an image is ready we can launch the multiple containers using that image.

 

Once you have your image ready, using Docker run you can launch your container. Docker run utility is a command that can create your entire OS install, boot up, and be ready to use within a second. Containers are designed to be transient and temporary.

 

Docker Engine is the core of Docker, the underlying client-server technology that creates and runs the containers. Generally speaking, when someone says Docker generically and isn’t talking about the company or the overall project, they mean Docker engine

 

The most important thing while learning Docker is running a container. Running a container is like launching our application so what will be the pre-requisites for that?

• Docker installed system
• An image

 

To check you have successfully installed the docker is or not, follow the command:

 

docker version

 

This will print the current version of Docker that you have installed.

 

 

Now the second pre-requisite is An image. I already told you about the online registry of the Docker named Docker Hub. To download the image from the public repository we have the command docker pull <image-name: version>. If you don't define the version it will download the latest version available.

 

I am going to install CentOs version 7

 

docker pull centos:7

 

I am going to do all my practicals on CentOs 7 and that's why I pulled the image of that Linux distro.

You can check your image is downloaded or not by listing all the images

 

docker images

 

Once, we have successfully downloaded the image we are ready to launch the container. Let's use the image that we have downloaded from the Docker Hub.

 

docker run -it --name myContainer centos:7

 

 

 

For running the container in the background we can add -dit option

 

docker run -dit --name myContainer centos:7

 

 

This will launch the container in the background and we will remain in the same shell.

 

You can list all your running container using the below command

 

docker ps

 

 

Using -a option you can list all the running and stopped command

 

docker ps -a

 

 

We can know details of our launched container using inspect command

 

docker inspect <container-name>

 

The above command will list all the details about the launched container.

 

 

You might need to do some configuration after launching the container and that's why we launch our container using -it option which provides us an interactive shell. Now, when you have completed your configuration you need to exit from this terminal, use a command

 

exit

 

This exit works fine, it will let you out from the container but it will also close the running container. If you want to close the container then there is no problem but in case you want to keep running your container and come out from it use ctrl+p+q.

 

This will don't close your container and safely land back to the bash terminal of your base application.

 

You can stop a running container using

 

docker stop container-name

 

This will stop your running container and you can easily check that after listing the running containers using docker ps.

Let's stop myContainer, which we have launched earlier 

 

 

Check the list of the running containers.

 

 

In order to restart your container again, use

 

docker start <container-name>

 

Let's  start myContainer 

 

 

 

Confirm using docker ps

 

 

 

How can you go inside your container again? We have known about the -it option which provides us the terminal of the container while launching it but what if we want to go back after coming out. Here is the command that can help you to do this task.

 

docker attach <container-name>

 

 

Yes, you can actually run a command inside the container without actually going inside it. There might be a case you want to run a command inside your container without going inside it. We have an option in Docker where we can actually run a command from our base OS and get the output in our base OS.

 

docker exec -it <container-name> <command>

 

Let's test this command in our launched container

 

docker exec -it myContainer whoami

 

whoami is a Linux command that provides the name of the user we logged in.

 

Using the above concept we can open the container in a better manner. How?

 

docker exec -it <container-name> bash

 

It will launch the container's bash shell for you. The better manner because after using command exit it will not close your running container. Don't forget to test in your container.

 

Logs are very essential while finding the error and fixing it. While working with a team it becomes very important to keep track of all the things. To find the log of any container follow the command -

 

docker logs <container-name>

 

Let's check the logs of our launched container 

 

docker logs myContainer

 

To know more about the handling of docker containers you can refer to the official docs name Docker docs. There you can search for your need and use the relevant command. You can use your base OS to know more about the command using the --help option after any command. For eg.

 

docker --help

 

Now, you have a very basic idea about docker and you also know how to run or stop a simple container. Let's discuss some core things.

 

Working inside the private world switches are essential to connect two or more containers. For the public world, we need routers to connect. 

 

Once, we launch a container docker-engine behind the scene create all of these. We don't have to worry about these complex configurations this is automatically done behind the scene and that is the reason we have connectivity between the containers and the public world. 

 

To list all the available networks you can use the below command

 

docker network ls

 

 

There you will find three different drivers - bridges, host, null. By default, the bridge driver is used while launching a container. These drivers are responsible for network connectivity.

 

Bridge Driver

Where driver bridge provides switch and routers both properties. Hence in case, you want connectivity with the outside world uses a bridge driver.

 

Host Driver

While the Host is used to copy all the networking settings of BaseOS, although it works only in Linux, Windows or Mac doesn't support this.

 

NULL Driver

The third driver NULL is used to keep disconnected the OS from the public world. If we don't want to connect our container from the public world we can choose this driver and it will keep our container private.

 

 

In order to create your own Docker network, you can follow the below command -

 

docker network create --driver bridge <network-name>

 

If you have more knowledge of networking there you can set the --subnet CIDR range and all. To know the details, use the --help option after creating for e.g. docker network create --help.

 

Once the network is created we can launch the container in it using the below command

 

docker run -it --name <container-name> --network <network-name> <image:version>

 

 

To overcome this, Docker has a facility that you can ping(connect) to the containers with their name rather than using their IP( they must be in the same network).

 

In case, if you launch two containers in the default network named bridge. And you want to ping with the name of the container you might fail. And if you try this setup in your custom network you can easily do this and that is the reason I already told you that custom network provides you some extra functionality.

 

But we have a trick to ping the container with the name inside the default network as well. To use this trick you can use the below command while launching the new container.

 

docker container run -it --link container1 --name <new-container> <image-name:version>

 

The above command will launch the container having a facility to ping container1(it is previously launched in the default network) with its name.

 

Now we are going to do a very great practical using the networking concept of Docker.

 

 

In this setup, we are going to set a web server and deploy our website there, and also set a load balancer to balance the traffic. We need to do the following setup

 

• A network with bridge driver.
• Two containers for our website having the same content.
• Configure and start the webserver.
• One more container as a client for accessing the website
• LoadBalancer to balance the load between both the pods.

 

In case, if you don't know about load balancing, It is a special type of layer over the containers and when the traffic comes up it first goes to the load balancer. The load balancer distributes the request over the containers equally. How does this could actually help?

 

Suppose, you have 3 containers running and suddenly one goes down client doesn't know about it and will never feel any latency in getting the data because the load balancer will connect the client with other running containers.

 

Launch a network

I already told you how you can launch your own custom network in my case I have used

 

docker network create --driver bridge webnet

 

 

We have created the network and named that webnet(you can do whatever you want). Check if it is successfully created or not by listing all the available networks inside your OS.

 

docker network ls

 

 

 

Launch two containers for web hosting

To host the website, I am going to use the Apache webserver. In the beginning, I told you in Docker we can download the images from the docker hub and also create one. Now, for this setup, I am going to use my previously created image.

 

You may use this or you can directly use the apache image. For the beginning it is advisable to use the above command because there might be some more configuration you have to do which I have already done in this image.

 

We need an image to launch so firstly pull the image(Downloading from Docker Hub).

 

docker pull anubhavsinghgtm/apache-php-webserver

 

 

Though while launching the container docker first check the local hard disk and then go to the public repository and pull the image from there. So, without manually pulling the image you can directly run the container it will first pull the image and then run the container. But it is always advisory to do the process step by step.

 

To set up the load balancer we are going to use the --network-alias option at the time of launching the container. For my setup, I have used the below command you can use on your own.

 

For container 1

docker run -dit --network webnet --name myc1 --network-alias website anubhavsinghgtm/apache-php-webserver

 

For container 2

docker run -dit --network webnet --name myc2 --network-alias website anubhavsinghgtm/apache-php-webserver

 

 

Here webnet is the name of our network which we have created earlier and network-alias is named as website. You can choose any name as per your choice.

 

List both the container is running or not using the docker ps command.

 

 

Ignore if you see this AH00558: httpd: Could not reliably determine the server's fully qualified domain name, using 172.18.0.3. Set the 'ServerName' directive globally to suppress this message.

 

Container as client

We need a client container to connect with the webserver in the same network. In this setup, we don't have public IP and therefore we can't access our webserver through the public world and that's the region our client should also be in the same network.

 

docker run -dit --network webnet --name client centos:7

 

 

Run docker ps. You will find three containers are running there.

 

Configure and start the webserver

If you are using the above-mentioned image you don't have to do anything but reading once to go through the complete process.

 

Go inside the container

 

docker exec -it myc1 bash

 

Inside the container go to /var/www/html (this is by default a directory for the webserver where we put all of our code). Create your PHP file there. If you are using the same image as I have used you will find an index.php file.

 

NOTE: If you are using another image it is more suggestible that create a PHP file where you can check the internal IP of your container and that will help us to understand the concept of load balancing.

 

You can use the below code to configure the index.php file

 

<pre>

<h2>Testing Apache Webserver over Docker</h2>

<?php print `ifconfig`?>

</pre>

 

NOTE: Make sure ifconfig command is running inside your container. If not? You can install the net-tools software. This will provide the ifconfig command for you and also install PHP to compile the above code.

(for centos, use the command yum install net-tools.)

 

Do the same steps in both of the containers.

 

Now, we are ready to check our results. Go inside the client container

 

docker exec -it client bash

 

 

Now you are inside your client container. First check you have connectivity with your container or not

 

ping -c 3 myc1

 

Similarly, for the myc2 container. Just replace myc1 with myc2.

 

Let's check our amazing setup. Use below command

 

curl website

 

 

You will see the IP of the containers. Try this 3-4 times and you will see sometimes you are connecting with your myc1 container and sometimes you are connecting with myc2 container.

 

Great!!! We have done great practical solving many use-cases of today's industries.

 

 

Inside your docker container if you want to connect with the public world you can do this easily. You can try this by the ping command.

 

ping -c 3 8.8.8.8

 

8.8.8.8 is one of Google's DNS IPs.

 

There are two types of connections we required - 

1.  Container to Public world: This process is done by SNATing, by default docker container has this property and therefore you would be able to connect to the internet.
2.  Public world to our container: This process is done by DNATing, by default docker container doesn't have this property. That means no buddy from the public can connect to our container.

 

Due to DNAting being disabled, in our previous setup, we have to launch our client in the same network webnet.

 

In the real world if we host our webserver we want to get clients from the public world. In that case, DNATing is very essential to be enabled. 

 

How to enable DNAT in the Docker container

 

We can't connect to the public IP from the private IP and similarly private IP from the public IP. So, to connect from the public IP to the Private IP we have to require a public IP and our container doesn't have any public IP.

 

Wait... What? If our container doesn't have any public IP and to connect to the public world we have also required a public IP, so how our container is able to connect? 

 

 

If this is the case, then the public world should also be able to connect with the container using the router's IP. But by default, this is disabled, why? Privacy is one of the main concerns and this makes the container more secure. 

 

Now for DNATing, we can connect the public world with the router's public IP and then write some rules for the router so that it replaces its public IP with the private IP of the container. There may be multiple services running inside one network therefore we use port numbers to easily redirect the request from the public client to the actual container.

 

To enable the DNATing we have to expose our container and to do so follow the below command -

 

docker run -it -p 8080:80 --name mywebserver httpd

 

By the above command, we have written a rule for the Port Address Translation, in which if there any request comes to the router from a particular port(8080) in our case, it redirects to the service running on port 80(used for HTTP).

 

That's it.

 

 

Generally, storage is of two types- Ephemeral and Persistent.

 

Ephemeral storage is that storage that loses the data once the OS is terminated. It is similar to the C drive of our windows.

 

Persistent storage doesn't lose its data even after the termination of the OS.

 

Docker containers are ephemeral in nature. If the container goes down(terminated), we can't get back our data. To overcome this issue docker has a concept of docker volume that can help the container to achieve persistent storage.

 

 

• Data would be safe even after the container terminated.
• Centralized storage: we can attach one volume to many different containers.
• Require less storage due to the centralized system.
• Easy to handle.

 

Add persistent storage in the container is like buying a hard disk and attaching that to our system. At the place of buying, we will create a volume and then add that volume to our container.

 

Create a volume

 

docker volume create <volume-name>

 

Add volume while launching

docker run -it --name mycontainer -v myvolume:/var/www/html anubhavsinghgtm/apache-php-webserver

 

While launching the webserver we want to keep the data related to webserver and that is the reason we have mounted the webserver directory to our newly created volume.  Your /var/www/html folder will be mounted with the myvolume, and a new container will be created.

 

···