The story of the development and deployment of any application in Docker begins with Docker images. But do you need to know how to create Docker images? Let's look at an example!
Suppose that you are a software developer and you want to develop an application that needs a Python runtime environment and some dependencies. But not just that, you need to run it on your cloud machine and have it work the same as it works on your local machine.
One way you might think of doing that is to download and install Python on your machine, configure the environmental variables, install all your application dependencies, and then run the application. Now, to run your application on the server, you need to do these steps all over again. To add a pinch of craziness to this situation, assume you want to scale that application and have to deploy that on hundred more servers. Even if you are a hard-working programmer, and you are ready to do it all, what if, while reviewing your application code, you discovered a bug or instead, you need to add a new feature? Bingo!!
Another, more efficient way, to deal with this type of problem is to create a Docker image of your application. For that, you need to grab a Python runtime as an image and then build your application alongside this Python runtime and all its dependencies. This is an easy and one-time process. But, once that image is ready, you can deploy your application on any Linux machine as many times you want. This application will get deployed in containers and is called containerization. Containers are flexible, lightweight, portable, and scalable.
I know you will now be excited to create your own first image but, before that, let us look at what Docker images and containers are. According to the documentation on Docker,
An image is an executable package that includes everything needed to run an application--the code, a runtime, libraries, environment variables, and configuration files.
A container is a runtime instance of an image--what the image becomes in memory when executed (that is, an image with the state, or a user process).
In this guide, we will explore and learn how to create Docker images and push it to the Docker Hub.
A Docker image can be compared to a git repository. Just like a git repository, it can be hosted on GitHub, Bitbucket, GitLab, or even a private git repo hosting service, but we could host our Docker image on Docker repository hosting service like Docker Hub.
Docker Hub is a service provided by Docker for hosting, finding, and sharing Docker Repositories. Just like git repo hosting services, a Docker repository can be public or private. A third-party repository hosting services also exists. The Docker Hub and other third party repository hosting services are called registries. For example, RedHat has their own registry to host their container images.
One important point to remember is that a registry has many repositories, while a repository has many different versions of the same image. These registries can be either private or public, depending on the needs of the organization. Docker Hub is one such example of a public registry.
Docker is configured to use Docker Hub as it's default registry. Use the
$ docker info command to view the registry that Docker is currently using. By default, it points to https://index.docker.io/v1/ which is the registry location for Docker Hub.
1$ docker info 2... 3Registry: https://index.docker.io/v1/ 4...
On Docker Hub, there are two kinds of images - official and unofficial. Official images are trusted and optimized. They have clear documentation, promote best practices, and are designed for the most common use cases. On the other hand, an unofficial image is any image that is created by a user. Docker Hub follows some standards so that both can easily be identified. Official images contain only the
<image_name> as its image name but unofficial images have the syntax as
<username>/<image_name>. Also, the official image has official written in the listing as shown in the below screenshot.
We could also look up and find images on Docker Hub by either using the search bar on the website or using the below command:
1$ docker search <image_name>
Let us search for an image with the name of busybox:
1$ docker search busybox 2NAME DESCRIPTION STARS OFFICIAL AUTOMATED 3busybox Busybox base image. 316 [OK] 4progrium/busybox 50 [OK] 5radial/busyboxplus Full-chain, Internet enabled ... 8 [OK] 6odise/busybox-python 2 [OK] 7azukiapp/busybox This image is meant to be used as ... 2 [OK]
To download a particular image, or set of images (i.e., a repository), use
docker pull. If no tag is provided, Docker Engine uses the
:latest tag as a default.
1$ docker pull <image_name>:<tag_name>
Let us pull the latest image of debian:
1$ docker pull debian 2Using default tag: latest 3latest: Pulling from library/debian 4fdd5d7827f33: Pull complete 5a3ed95caeb02: Pull complete 6Digest: sha256:e7d38b3517548a1c71e41bffe9c8ae6d6d29546ce46bf62159837aad072c90aa 7Status: Downloaded newer image for debian:latest
We could create our own Docker image in two ways:
A Dockerfile is a simple text document that contains a series of commands that Docker uses to build an image. Several commands supported in Dockerfile are FROM, CMD, ENTRYPOINT, VOLUME, ENV, and more. A simple Dockerfile looks as follows:
1FROM busybox:latest 2CMD ["date"]
Note: An important point to remember is that this file should be named as Dockerfile.
docker build command builds an image from a Dockerfile. To build the image from our above Dockerfile, use this command:
1$ docker build -t example_image . 2Sending build context to Docker daemon 2.048kB 3Step 1/2 : FROM busybox:latest 4---> db8ee88ad75f 5Step 2/2 : CMD ["date"] 6---> Using cache 7---> b89335fdacdf 8Successfully built b89335fdacdf 9Successfully tagged example-image:latest
Now, let’s run a container based on our image. You will find that it will print out the date as shown below:
1$ docker run -it --name example_app example_image 2Sun Jul 21 20:52:20 UTC 2019
Another way to create an image is by pulling a Docker image, creating a container from it, and then modifying or making changes in it like installing our app in that container. Now, using the
docker commit command, we can create a Docker image from the container.
Let's look at an example of how to create a Docker image from our example_app container:
1$ docker ps -a 2CONTAINER ID IMAGE COMMAND CREATED STATUS NAMES 3c3df2dd33276 example-image "date" 5 seconds ago Exited (0) 4 seconds ago example_app 4 5$ docker commit example_app example_image2:latest 6sha256:7b48e8355aa7a7ea32d554f26d0bd21f4d069d8526c68f1d098acac9111a9adf
To be able to publish our Docker images to Docker Hub, there are some steps that we need to follow:
Before we can push our image to Docker Hub, we will first need to have an account on Docker Hub. Create an account by visiting this link. The signup process is relatively simple.
For uploading our image to Docker Hub, we first need to create a repository. To create a repo:
example-image, the Docker image that we created earlier using Dockerfile. Also, describe your repo like "My First Repository". Finally, click on the create button. Refer to the below screenshot:
Now we will push our built image to the Docker Hub registry:
Log into the Docker public registry from your local machine terminal using Docker CLI:
1$ docker login
Tag the image
This is a crucial step that is required before we can upload our image to the repository. As we discussed earlier, Docker follows the naming convention to identify unofficial images. What we are creating is an unofficial image. Hence, it should follow that syntax. According to that naming convention, the unofficial image name should be named as follows:
<username>/<image_name>:<tag_name>. In my case, I need to rename it as
1$ docker tag example_image:latest gauravvv/example_image:latest
Publish the image
1$ docker push gauravvv/example_image:latest
Upload your tagged image to the repository using the
docker push command. Once complete, you can see the image there on Docker Hub. That's it; you have successfully published your Docker image. If you want to test out your image, use the below command and launch a container from it:
1$ docker pull gauravvv/example_image:latest 2$ docker run -it gauravvv/example_image:latest
So far, we have learned how to create a new Docker image, tag it, and push it to the Docker Hub. While using Kubernetes or OpenShift, we would use the internally hosted Docker registry; some organizations also prefer to host their own internal private registry. Therefore, you should know how to publish your Docker image to an internally hosted Docker registry. Luckily, Docker made it more accessible, just like publishing to Docker Hub but a little bit different.
docker login command to log in to the specified registry and to tell Docker which registry we want to sign in to.
1$ docker login registry.example.com
As you may remember, for publishing our image to Docker Hub, we have tagged the image to include the username at first. Now, to publish to the internal registry, continuing on that pattern, we will also add the registry name at first, followed by a username, then followed by image name. Therefore, the final syntax will look like this:
<registry>/<username>/<image_name>:<tag>. Let's take a look at an example:
1$ docker tag gauravvv/example_image registry.example.com/gauravvv/example_image
docker push command takes the name of the image. It will know where to push our Docker image by looking at the image name because the name contains the registry location. By looking at our image name, which is
registry.example.com/gauravvv/example_image , it will determine that we are pushing an image name,
example_image, to a user's repository,
gauravvv, at the
1$ docker push registry.example.com/gauravvv/example_image
Congratulations on finishing up this guide! I hope you have enjoyed learning how to create a Docker image and publish it to Docker Hub. Understanding the use cases of Docker is as vital as learning its concepts. Now, go ahead, Dockerize your applications and share them with the world. Happy Dockerizing!