Chapter 9. Operating Systems, Platforms, and Cloud and Local Providers

The first half of this chapter will cover how open standards encourage a diverse ecosystem of container implementations. We'll look at the Open Container Initiative (OCI) and its mission to provide an open container specification as well. The second half of this chapter will cover the various operating systems available for running containerized workloads, such as CoreOS. We'll also look at its advantages as a host OS, including performance and support for various container implementations. Additionally, we'll take a brief look at the Tectonic Enterprise offering from CoreOS. We'll look at the various hosted platforms offered by the major cloud service providers (CSPs) and see how they stack up.

This chapter will discuss the following topics:

• Why do standards matter?
• The OCI and the Cloud Native Computing Foundation (CNCF)
• Container specifications versus implementations
• Various container-oriented operating systems
• Tectonic
• The...

Technical requirements

You'll need to have your Google Cloud Platform account enabled and logged in, or you can use a local Minikube instance of Kubernetes. You can also use Play with Kubernetes online at https://labs.play-with-k8s.com/.

You'll also need GitHub credentials, which we'll go over setting up later in the chapter.

 

The GitHub repository for this chapter can be found at https://github.com/PacktPublishing/Getting-Started-with-Kubernetes-third-edition/tree/master/Code-files/Chapter09.

The importance of standards

Over the past two years, containerization technology has had a tremendous growth in popularity. While Docker has been at the center of this ecosystem, there is an increasing number of players in the container space. There are already a number of alternatives to the containerization and Docker implementation itself (rkt, Garden, and so on). In addition, there is a rich ecosystem of third-party tools that enhance and complement your container infrastructure. While Kubernetes is designed to manage the state of a container and the orchestration, scheduling, and networking side of this ecosystem, the bottom line is that all of these tools form the basis to build cloud-native applications.

As we mentioned at the very beginning of this book, one of the most attractive things about containers is their ability to package our application for deployment across various environment tiers (that is, development, testing, and production) and various infrastructure providers (GCP...

The OCI

One of the first initiatives to gain widespread industry engagement is the OCI. Among the 36 industry collaborators are Docker, Red Hat, VMware, IBM, Google, and AWS, as listed on the OCI website at https://www.opencontainers.org/.

The purpose of the OCI is to split implementations, such as Docker and rkt, from a standard specification for the format and runtime of containerized workloads. According to their own terms, the goal of the OCI specifications has three basic tenets (you can refer to more details about this in the Further reading section at the end of the chapter):

• Creating a formal specification for container image formats and runtime, which will allow a compliant container to be portable across all major, compliant operating systems and platforms without artificial technical barriers.
• Accepting, maintaining, and advancing the projects associated with these standards. It will look to agree on a standard set of container actions (start, exec, pause, and so on), as well as...

CNCF

A second initiative that also has widespread industry acceptance is the CNCF. While still focused on containerized workloads, the CNCF operates a bit higher up the stack, at the application design level.

 

 

Its purpose is to provide a standard set of tools and technologies to build, operate, and orchestrate cloud-native application stacks. Cloud has given us access to a variety of new technologies and practices that can improve and evolve our classic software designs. The CNCF is also particularly focused on the new paradigm of microservice-oriented development.

As a founding participant in the CNCF, Google has donated the Kubernetes open source project. The goal will be to increase interoperability in the ecosystem and support better integration with projects. The CNCF already hosts a variety of projects on orchestration, logging, monitoring, tracing, and application resilience.

Note

For more information on CNCF, refer to https://cncf.io/.

We'll talk more about the CNCF, Special Interest...

Standard container specification

A core result of the OCI effort is the creation and development of the overarching container specification. The specification has five core principles that all containers should follow, which I will briefly paraphrase:

• The container must have standard operations to create, start, and stop containers across all implementations.
• The container must be content-agnostic, which means that type of application inside the container does not alter the standard operations or publishing of the container itself.
• The container must be infrastructure-agnostic as well. Portability is paramount; therefore, the container must be able to operate just as easily in GCE as in your company's data center or on a developer's laptop.

• A container must also be designed for automation, which allows us to automate across the build, as well as for updates and the deployment pipelines. While this rule is a bit vague, the container implementation should not require onerous manual steps for creation...

CoreOS

While the specifications provide us with a common ground, there are also some trends evolving around the choice of OS for our containers. There are several tailored-fit OSes that are being developed specifically to run container workloads. Although implementations vary, they all have similar characteristics. The focus is on a slim installation base, atomic OS updating, and signed applications for efficient and secure operations.

 

 

One OS that is gaining popularity is CoreOS. CoreOS offers major benefits for both security and resource utilization. It provides resource utilization by completely removing package dependencies from the picture. Instead, CoreOS runs all applications and services in containers. By providing only a small set of services required to support running containers and bypassing the need for hypervisor usage, CoreOS lets us use a larger portion of the resource pool to run our containerized applications. This allows users to gain higher performance from their infrastructure...

Kubernetes with CoreOS

Now that we understand the benefits, let's take a look at a Kubernetes cluster using CoreOS. The documentation supports a number of platforms, but one of the easiest to spin up is AWS with the CoreOS CloudFormation and CLI scripts.

Note

If you are interested in running Kubernetes with CoreOS on other platforms, you can find more details in the CoreOS documentation at https://coreos.com/kubernetes/docs/latest/. You can find the latest instructions for AWS athttps://coreos.com/kubernetes/docs/latest/kubernetes-on-aws.html.

You can follow the instructions covered previously in this chapter to spin up Kubernetes on CoreOS. You'll need to create a key pair on AWS, and also specify a region, cluster name, cluster size, and DNS to proceed.

In addition, we will need to create a DNS entry, and will require a service such as Route 53 or a production DNS service. When following the instructions, you'll want to set the DNS to a domain or sub-domain on which you have permission to...

Tectonic

Running Kubernetes on CoreOS is a great start, but you may find that you want a higher level of support. Enter Tectonic, the CoreOS enterprise offering for running Kubernetes with CoreOS. Tectonic uses many of the components we already discussed. Both Docker and rkt runtimes are supported. In addition, Kubernetes, etcd, and flannel are packaged together to give a full stack of cluster orchestration. We discussed flannel briefly in Chapter 3, Working with Networking, Load Balancers, and Ingress. It is an overlay network that uses a model similar to the native Kubernetes model, and uses etcd as a backend.

Offering a support package similar to Red Hat, CoreOS also provides 24/7 support for the open source software that Tectonic is built on. Tectonic also provides regular cluster updates and a nice dashboard with views for all of the components of Kubernetes. CoreUpdate allows users to have more control of the automatic update process. In addition, it ships with modules for monitoring...

Hosted platforms

There are several options available for hosted Kubernetes in the cloud. These Platforms as a service (PaaS) can provide a stable operating model as you push towards production. Here's an overview of the major PaaSes provided by Amazon, Microsoft, and Google.

Amazon Web Services

Elastic Container Service (ECS) has just been launched as of the time of this chapter's writing. AWS is preparing a networking plugin to differentiate itself from other offerings, called the vpc-cni. This allows for pod networking in Kubernetes to use Elastic Network Interfaces (ENIs) on AWS. With ECS, you do have to pay for manager nodes, which is a different path to that taken by Microsoft and Google. ECS' startup procedure is also currently more complex and doesn't have single-command creation via the CLI.

Microsoft Azure

The Azure Container Service is the second longest running hosted Kubernetes service in the cloud after the Google Kubernetes Engine. You can use Azure templates and the Resource Manager...

Summary

In this chapter, we looked at the emerging standards bodies in the container community and how they are using open specifications to shape the technology for the better. We looked at various container frameworks and runtimes. We dipped our toes into the CNCF, and tried out CRI-O.

We also took a closer look at CoreOS, a key player in both the container and Kubernetes community. We explored the technology that CoreOS is developing in order to enhance and complement container orchestration, and saw first-hand how to use some of it with Kubernetes. Finally, we looked at the supported enterprise offering of Tectonic and some of the features that are available now.

We also looked at some of the major PaaS offered by cloud service providers.

In the next chapter, we will explore the broader Kubernetes ecosystem and the tools available to move your cluster from development and testing into full-blown production.

Further reading

• https://www.opencontainers.org/faq/(underHow broad is the mission of the OCI?)
• https://github.com/opencontainers/specs/blob/master/principles.md