For Mac and Windows the installation could not be simpler. All you need to do is download and install the Docker Toolbox found at https://www.docker.com/toolbox. The installer includes the Docker Client, Docker Machine, Compose (Mac only), Kitematic, and VirtualBox.
Since Docker is based on the Linux Container technologies which are not available on Mac and Windows, VirtualBox is used to run a tiny Linux kernel containing the Docker server.
At the time of this writing installing Docker on Linux is not as easy. To install Docker on Linux you may have to install some prerequisites; check https://docs.docker.com/installation for specific instructions. For some distributions there may be packages available using its native package manager. For other distributions you will need to run:
curl -sSL https://get.docker.com/ | sh
Optionally on Linux you can install Docker-Machine as root; to do so, execute the following:
curl -L https://github.com/docker/machine/releases/download/v0.4.0/docker-machine_linux-amd64 > /usr/local/bin/docker-machine
chmod +x /usr/local/bin/docker-machine
If you want to create machines locally, you will also need to install VirtualBox using the instructions found at https://www.virtualbox.org/wiki/Linux_Downloads.
As of the date of this publication, Docker-Machine is still considered in beta and is not recommended for production use.
Running a Container
With Docker installed you are able to begin running containers. If you don’t already have the container you want to run, Docker will download the image necessary to build the container from the Docker Hub, then build and run it.
To run the simple hello-world container to make sure everything is configured properly, run the following commands:
docker run hello-world
Ultimately this command prints a message to standard output explaining all the steps that took place to display the message.
Typical Local Workflow
Docker has a typical workflow that enables you to create images, pull images, publish images, and run containers.
The typical Docker workflow involves building an image from a Dockerfile, which has instructions on how to configure a container or pull an image from a Docker Registry such as Docker Hub. With the image in your Docker environment, you are able to run the image, which creates a container as a runtime environment with the operating systems, software, and configurations described by the image. For example, your result could be a container on the Debian operating system running a version of MySQL 5.5, which creates a specific database with users and tables required by your web application. These runnable containers can be started and stopped like starting and stopping a virtual machine or computer. If manual configurations or software installations are made, a container can then be committed to make a new image that can be later used to create containers from it. Finally, when you want to share an image with your team or the world, you can push your images to a Docker registry.
Pull Image From Docker Registry
The easiest way to get an image is to visit https://hub.docker.com and find an already prepared image to build a container from. There are are many certified official accounts for common software such as MySQL, Node.js, Java, Nginx, or WordPress, but there are also hundreds of thousands of images created by ordinary people as well. If you find an image you want, such as mysql, execute the pull command to download the image.
docker pull mysql
If you don’t already have the image locally, Docker will download the most current version of that image from Docker Hub and cache the image locally. If you don’t want the current image and instead want a specific version, you can also a tag to identified the desired version.
docker pull mysql:5.5.45
If you know you will want to run the image immediately after pulling, you can save a step by just using the run command and it will automatically pull it in the background.
Building Image From a Dockerfile
If you can’t find what you need or don’t trust the source of an image you find on Docker Hub, you can always create your own images by creating a Dockerfile. Dockerfiles contain instructions for inheriting from an existing image, where you can then add software or customize configurations.
The following is a simple example of what you might find in a file named Dockerfile:
RUN echo America/New_York | tee /etc/timezone && dpkg-reconfigure --frontend noninteractive tzdata
This Dockerfile example shows that the image created will inherit from the certified mysql repository (specifically the 5.5.45 version of MySQL). It then runs a Linux command to update the time zone to be Eastern Time.
More details on creating a Dockerfile will be provided later.
To build this image from the directory containing the Dockerfile, run the following command:
docker build .
This command will create an unnamed image. You can see it by running the command to list images.
This displays all the locally cached images, including the ones created using the build command.
REPOSITORY TAG IMAGE ID VIRTUAL SIZE
<none> <none> 4b9b8b27fb42 214.4 MB
mysql 5.5.45 0da0b10c6fd8 213.5 MB
As you can see, the build command created an image with a repository name and tag name of <none>. This tends not to be very helpful, so you can use a –t option to name the image for easier usage:
docker build –t est-mysql .
Listing the images again you can see the image is much clearer.
REPOSITORY TAG IMAGE ID VIRTUAL SIZE
est-mysql latest 4b9b8b27fb42 214.4 MB
mysql 5.5.45 0da0b10c6fd8 213.5 MB
There is an alternative option to creating a custom image besides writing a Dockerfile. You can run an existing image with bash access then customize the image manually by installing software or changing configurations. When complete you can run the docker commit command to create an image of the running container. This is not considered a best practice since it is not repeatable or self documenting like using the Dockerfile method.
Running an Image
To run a Docker image you just need to use the docker run command followed by a local image name or one found in Docker Hub. Commonly, though, a Docker image will require some additional environment variables, which can be specified with the -e option. For long running process like daemons you also need to use a –d option. To start the est-mysql image, you would run the following command to configure the MySQL root user’s password, as documented in the Docker Hub mysql repository documentation:
docker run -e +1 -d est-mysql
To see the running container, you can use the Docker ps command:
The ps command lists all the running processes, the image name they were created from, the command that was run, any ports that software are listening on, and the name of the container.
CONTAINER ID IMAGE COMMAND PORTS NAMES
30645f307114 est-mysql "/entrypoint.sh mysql" 3306/tcp serene_brahmagupta
As you can see from the running processes above, the name of the container is serene_brahmagupta. This is an auto-generated name and may be challenging to maintain. So it is considered a best practice to explicitly name the container using the --name option to provide your name at container start up:
docker run --name my-est-mysql -e +1 -d est-mysql
You will notice from the ps output that the container is listening to port 3306, but that does not mean you are able to use the mysql command line or MySQL Workbench locally to interact with the database, as that port is only accessible in the secure Docker environment in which it was launched. To make it available outside that environment, you must map ports using the –p option.
docker run --name my-est-mysql -e +1 -p 3306:3306 -d est-mysql
Now mysql is listening on a port that you can connect to. But you still must know what the IP address is to connect. To determine the IP address you can use the docker-machine ip command to figure it out.
docker-machine ip default
Using default as the machine name, which is the default machine installed with the Docker Toolbox, you will receive the IP address of the machine hosting your docker container.
With the IP address, you can now connect to mysql using your local mysql command line.
mysql -h 192.168.99.100 -u root -proot+1
Stopping and Starting Containers
Now that you have a Docker container running, you can stop it by using the Docker stop command and the container name:
docker stop my-est-mysql
The entire state of the container is written to disk, so if you want to run it again in the state it was in when you shut it down, you can use the start command:
docker start my-est-mysql
Tagging an Image
Now that you have an image that you have run and validated, it is a good idea to tag it with a username, image name, and version number before pushing it to repository. You can accomplish this by using the Docker tag command:
docker tag est-mysql javajudd/est-mysql:1.0
Push Image to Repository
Finally, you are ready to push your image to Docker Hub for the world to use or your team to use via a private repository. First if you haven’t done so already you will need to go https://hub.docker.com/ to create a free account. Next you need to login using the docker login command.
When prompted, input the username, password, and email address you registered with.
Now push your image using the push command, specifying your username, image name, and version number.
docker push javajudd/est-mysql:1.0
After some time you will receive a message that the repository has been successfully pushed. If you log back into your Docker Hub account, you will see the new repository.