Container Terminology

As with any new technology, it is important to understand the terminology that goes along with that new technology. The first thing that you may have noticed is that a container works a lot like a virtual machine. Just like a virtual machine, the container has a running operating system within the container.

The container has a file system and the container can also be accessed through the network the same way a virtual machine does. The advantage is that a container is a more efficient operating system. But to truly understand how containers work, you need to understand all of the different components that allow containers to function properly.

Container Host This component can be on a physical or virtual machine and it’s the component that is configured with the Windows Container feature. So the Windows Container sits on top of the Container Host.

Container OS Image This component provides the operating system to the container. Containers are made up from multiple images that are stacked on top of each other within the container.

Container Image This is the component that contains all of the layers of the container. So the Container Image contains the operating system, the application, and all of the services required to make that application function properly.

Container Registry This component is the heart and brain of the container. The container images are kept within the container’s registry. The advantage of doing containers this way is that you can download other registries to automatically add other applications or services quickly.

Docker Daemon This is the component that runs the docker application. The docker daemon is automatically installed after you complete the installation of the docker application. If you need to configure the docker daemon, you would use the docker daemon file. This file is in a JSON format.

Dockerfile This component is used to create the container images. The advantage of using the Dockerfile is that you can automate how containers are created. Dockerfiles are batches of instructions (within a txt file) and commands that are called on when an image is assembled.

Docker Hub Repositories This component is a location where all of your images are stored. By having a central location for stored images, the images can be used among co-workers, customers, or the for the entire IT community. There are docker hub repositories on the internet and these locations allow you to grab and use images for your organization.

Install and Configure Server Containers

So now that we have talked about what the different components are of a container, it’s time that we look at installing containers on our Windows Server 2016 system. When it comes to Microsoft, it really doesn’t matter if we are installing containers on a GUI based system or non-GUI based system. We are going to install the components needed by using Windows PowerShell.

But before we can look at installing and using containers, I need to show you what is required on the Windows Server 2016 system. So the first step in using containers is looking at what we need on our network and computers for containers to run properly.

Supported Images for Windows Containers

When I talk about setting up containers and you are getting ready to start using containers, there is one major requirement that we need to consider. The operating system on the host machine needs to be the same operating system that is used in the Windows Container. If you install a different operating system in the Windows Container, the container may load but you will most likely start to see errors and there is no guarantee that all of the container’s functionality will work.

So it is very important to make sure that the version of Windows Server 2016 that you install onto the host system is the same version as you run in the Windows Container. One nice advantage to using Microsoft Windows is that you can check what version of Windows you are using.

To see what Windows version you have installed, enter into the system’s Registry (Regedit.exe) and search the following Registry key (see Figure 7.1):

HKEY_LOCAL_MACHINESoftwareMicrosoftWindows NTCurrentVersion

Depending on which host operating system that you are going to run will determine what operating systems you can run in the Windows Server Container or Hyper-V container. Not all operating systems are available depending on the host OS image. Table 7.1 shows you all of the supported configurations for each host operating system.

Host Operating System	Windows Server Container	Hyper-V Container
Windows Server 2016 with Desktop	Server Core / Nano Server	Server Core / Nano Server
Windows Server 2016 Core	Server Core / Nano Server	Server Core / Nano Server
Nano Server	Nano Server	Server Core / Nano Server
Windows 10 Pro / Enterprise	Not Available	Server Core / Nano Server

Install and Configure Windows Containers

So now that you have installed docker on your Windows Server 2016 system, let’s take a look at how you install and configure containers.

In Exercise 7.2, I will show you how to make sure your Docker service is started. I will then show you how to install a base operating system image onto your host. I will also show you how to create Windows Server Containers.

Now that you have seen how to download an image to the systems repository, you need to see how you can turn an image into a container. To do this, you need to just use the docker run command to get the image into a container. Also there may be a time when you need to remove an image from a container. This is done using the docker rm command.

But first you need to know which image that you want to put into a container. Exercise 7.3 will show you how to see your images and then how to add an image into a container. To complete Exercise 7.3, you must have completed Exercise 7.2.

Tagging an Image

Administrators have the ability to change the tags associated with the images. Many administrators use tag names as version names so that they can keep track of the different images on their machine.

By having tags that you create, it allows you to easily access the images later by its tag name. To tag an image, you use the -t parameter. So the following is tagging an image as WillPanekImage.

docker build -t WillPanekImage

Uninstall an Operating System Image

One of the maintenance issues that many IT administrators must deal with is hard drive space. As you are downloading and using images, there may be a time when you need to remove images from our server that are no longer being used.

If you continue to download and use multiple images files, you will need to make sure that every once in a while that you look at all of your images and delete any images that are no longer valid or no longer in use.

Exercise 7.4 will show you how to remove an image file from your host system. To complete this exercise, make sure that you have completed the earlier exercises in this chapter.

Creating New Images Using Dockerfile

One nice advantage of using docker is that you can go out to Docker’s website and look at the different images that are available. There are images for operating systems, applications, and software.

But what if you can’t find an image that you need? Well, then you can build your own images using Dockerfiles. When an administrator installs docker, the docker engine includes tools that the IT department can use to create these Dockerfiles. Dockerfiles are just text files that are manually created and they are compiled and turned into an image file.

If your organization decides that they want to build their own Dockerfiles, then they will get some benefits while doing just that. Some of the advantages of building your own Dockerfiles are as follows:

• Administrators can store images as code.
• Your organization can have rapid recreation of images. These images can then be used for maintenance and upgrade cycles.
• Customization for exactly what your organization wants.

So once you have installed docker, there will be some components that are included so that you can create your own Dockerfiles. These two docker components include the docker engine and the compiler (docker build command).

So if you have ever built .ini files or even a host file, then you understand how building a file can work. Same as many programming codes or .ini files, you can use the pound (#) to show comments in the file. This is very useful.

Many years ago, before I got into networking, I was a programmer. One thing that most programmers hate is when you look at someone else’s coding and you have no idea what they were doing. When a coder takes the time to put in comments so that anyone can follow them and work on the code, it makes following that coder a thousand times easier. This is what the comments in the Dockerfile do. Using the pound (#) for making comments and stating exactly why each line is included, helps someone following you or if someone is trying to learn what you do.

So let’s take a look at an example of a Dockerfile:

# Sample Dockerfile for WillPanek
# We will be using Windows Server Core as our base image.
FROM microsoft/windowsservercore
# Uses dism.exe to install the DNS role.
RUN dism.exe /online /enable-feature /all /featurename: : DNS-Server-Full-Role /NoRestart
# Sets a command or process that will run each time a container is run from the new image.
CMD [ "cmd" ]

So now that we can see a Dockerfile, lets break down some of the different sections that you can configure. Table 7.4 will show you some of the configuration settings that you can use.

Understanding Hyper-V Containers

So far in this chapter we have discussed Windows Containers, but now we are going to look at Hyper-V Containers. As I stated earlier, Windows Containers share the system’s kernel between all containers and the host. Hyper-V Containers are different because each Hyper-V Containers utilizes their own instance of the Windows kernel.

Since Hyper-V containers utilizing their own instance of the Windows kernel, administrators can use different versions of Windows between the host system and the image version.

Also the Windows host system needs to have the Microsoft Hyper-V role installed onto the host system. Windows Server 2016 and Windows 10 Professional and Enterprise (Anniversary Editions) both allow you to create containers and Hyper-V.

The one nice feature is that both container types, Windows Containers and Hyper-V Containers, are created, managed, and function the exact same way. The only difference is how the Hyper-V Containers have better isolation to the kernel.

When you are working with Hyper-V containers in Docker, the settings are identical to managing Windows Server containers. The one difference that you want to include in the Hyper-V Container is using the ––isolation=hyperv parameter. The following is an example of the docker command with the Hyper-V parameters:

docker run -it ––isolation=hyperv microsoft/nanoserver cmd

Managing Container Networking

One of the features included with building containers is the ability to access the servers and data within the container the same way you would on a normal network server or Hyper-V server.

Once you have installed Docker, there will be three networks that are created automatically. You can see these networks by typing docker network ls in PowerShell (see Figure 7.13) or at an elevated command prompt.

If you would like to get even more details about a specific network (see Figure 7.14), after you run the docker network ls command, you will need to grab the Network ID number. Then type the following PowerShell command followed by the Network ID number (my Network ID is 3cb894810795):

Docker network inspect 3cb894810795

One nice thing about working with networks within containers is that these three networks are always available to you even when you choose only one to be part of your container. You can specify which network you want your container to run on by using the ––Network flag setting.

When you create a container, the host network adds the container onto the host’s network stack. When you are dealing with the containers network, there are very few reasons why you would even need to manage or manipulate the container’s network. The only network that you may need to work with is the bridge network. The docker default bridge is created as soon as you install the Docker Engine. It creates your bridge network and its name is bridge.

Using Docker Hub Repository

One really nice advantage of using containers is that there are hundreds of images that you can use. Docker has a public database of images that you can access.

The DockerHub repository has images for Microsoft, UNIX, Linux, and hundreds more. If you want to see what a vendor has out on the repository, just type in docker search vendorname. So let us take a look at what Microsoft has for you in the repository. Type docker search Microsoft (see Figure 7.15).

Administrators have the ability to setup a private repository so that co-workers can share and use the images that you create. After you create your images using the Docker Daemon, you can then push those images to your corporate Docker Hub repository. Administrators can add users and accounts to the Docker Hub to verify that only the organization’s users are accessing the images.

If you are building images and placing those images on GitHub or Bitbucket, you can then use the Automatic build repository that is included with the Docker Hub service.

So when you are ready to start uploading corporate images to the Docker Hub, you will first need to create a Docker Hub user account (https://cloud.docker.com/). After you have created your account, click the Create ↓ and choose “Create Repository.”

You will then be asked to put in a Docker ID namespace for your organization. The repository name needs to be unique and it can be up to 255 characters. The namespace will only allow letters, numbers, or the dash (-) and underscore (_). You are then allowed to put in a Short Description (100 characters or less) and a Full Description. Then you click the Create button and you are finished.

After your repository is created, you can push images to the repository by putting in the name of your image, your Docker Hub username, the repository name that you created earlier, and the image tag. The following is an example of the docker push command:

docker push <hub-user>/<repo-name>:<tag>

Using Microsoft Azure for Images

If you decide to run container applications within a cluster of virtual machines, you can use the Azure Container Service. The Azure Container Service allows you to easily create, configure, and manage your virtual machine cluster of containers. The Azure Container Service uses open-source scheduling and management tools.

By using open-source tools, the Azure Container Service connects you with thousands of other users who are also designing, building, and maintaining container images.

The Azure Container Service uses the Docker format but it is also compatible with Marathon, DC/OS, Kubernetes, or Docker Swarm. Because the Azure Container Service works with all of these different formats, you have the ability to work with thousands of applications and images. But since this is Microsoft’s Azure platform, you get all of the security benefits and features that Azure has to offer.

To setup the Azure Container Service, you must first setup an Azure Container Service cluster through the Microsoft portal. Once you have entered the portal, you would need to use the Azure Resource Manager template for Docker Swarm, DC/OS, and Kubernetes, or you would need to use the Azure Command Line Interface.