1: Setting the scene

This chapter doesn’t map directly to an exam objective. However, the things you’ll learn are in the exam and are important in setting the scene for why we have technologies like containers and Kubernetes. If you already know this, you can skip to the next chapter. Otherwise, stick around while we set the scene for the rest of the book.

We’ll cover all of the following at a high level.

• Virtualisation
• Containerisation
• The transition from monolithic apps to microservices

Don’t worry if you think we’re not covering things in enough detail. This is just an introductory chapter and we’ll get into the detail in later chapters.

Virtualisation

In the distant past we deployed one application per physical server. This was a huge waste of physical resources and company capital. It also delayed the rollout of applications while physical servers were procured, racked, patched into the network, and had an operating system installed.

Virtualisation technologies like VMware came along and opened the door for us to run multiple applications on a single physical server. This meant we didn’t have to buy a new server for every new app, we could deploy apps very quickly to virtual machines on existing servers and avoid all of the following:

• No more waiting for finance to approve server purchases
• No more waiting for the datacenter team to rack and cable servers
• No more waiting for the network team to authorise servers on the network
• No more waiting for sysadmins to install operating systems

Almost immediately we went from wasting money on over-powered physical servers that took ages to purchase and install...

Containerisation

In the early 2010’s Docker gave the world the gift of easy-to-use containers.

At a high level, containers are another form of virtualisation that allow us to run even more apps on less servers and deploy them even faster.

Figure 1.1 shows a side-by-side comparison of server virtualisation and container virtualisation.

Figure 1.1

As the image shows, server virtualisation slices a physical server into multiple virtual machines (VM). Each VM looks, smells, and feels like a physical server, meaning each one has virtual CPUs, virtual memory, virtual hard drives, and virtual network cards. You install an operating system (OS) on each one and then install one app per VM. If a single physical server is sliced into 10 virtual machines, there will be 10 operating systems and you can install 10 apps.

Container virtualisation slices operating systems into virtual operating systems called containers. Each container looks, smells, and feels like a normal...

Monolithic vs microservices

In the past, we built monolithic applications that ran on dedicated physical servers.

Monolithic application is jargon for a large complicated application that does lots of things. For example, a monolithic application may have all of the following services bundled and shipped as a single program.

• Web front-end
• Authentication
• Shopping basket
• Catalog
• Persistent store
• Reporting

The important thing to understand is that all of these services were developed by a single team, shipped as a single program, installed as a single program, and patched and updated as a single program. This meant they were complex and difficult to work with. For example, patching, updating, or scaling the reporting service of a monolithic app meant you had to patch, update, or scale the entire app. This made almost all changes high-risk, often resulting in updates being rolled up into a single very high-risk update performed once a year over a long stressful weekend...

Chapter summary

In this chapter, you learned that we used to deploy one application per physical server. This was wasteful of capital, servers, and environmental resources. It also caused long delays in application rollouts while new servers had to be procured, delivered, and installed. VMware came along and let us run multiple applications per physical server. It reduced capital expenditure and allowed more efficient use of server and environmental resources. It also allowed us to ship applications a lot faster by deploying them to virtual machines on servers we already owned.

Containers are also a form of virtualisation. They virtualise at the operating system layer and each container is a virtual operating system. Containers are faster and more efficient than virtual machines, however, out-of-the-box they’re usually less secure. The advantages of containers made it possible for us to re-think the way we develop, deploy, and manage applications.

A major innovation, enabled...

Exam essentials

This chapter doesn’t map directly to an exam domain. However, the following exam topics were mentioned and will be covered in more detail later in the book.

Container runtimes

A server that runs containers is called a container host. They use a low-level tool called a container runtime to start and stop containers. Docker is the best-known container runtime and was the first container runtime supported by Kubernetes. However, it is being replaced in Kubernetes by a lighter-weight version called containerd (pronounced “container dee”). Many other container runtimes exist and some of them work differently. Some offer better performance at the expense of security, whereas others offer better security at the expense of size and performance. You’ll learn more later in the book.

Container security

All containers running on a single host share the host’s OS kernel. This makes them small, portable, and fast to start. However, if the...

Recap questions

See Appendix A for answers. Page 157 in the paperback edition.

1. Which of the following are advantages of containers compared to virtual machines? Choose all correct answers.

• 1. Smaller size
• 2. Faster start times
• 3. More secure out-of-the-box
• 4. More apps per physical server

2. Which of the following is an advantage of virtual machines compared to containers?

• 1. Virtual machines start faster than containers
• 2. Virtual machines are more secure out-of-the-box
• 3. Virtual machines are smaller than containers
• 4. Virtual machines enable microservices design patterns

3. Which layer does container virtualisation work at?

• 1. The hardware layer
• 2. The infrastructure layer
• 3. The application layer
• 4. The operating system layer

4. Which of the...