Understanding Kubernetes at enterprise scale

When Kubernetes first emerged, many developers were excited to get their hands on this powerful new containerization technology. In many cases, container usage has spread from the bottom up as developers experiment and learn how to leverage it to its full potential. Today, there’s no question container adoption is exploding. Gartner estimates that more than 75% of global organizations will be running containerized applications in production by 2022, compared to 30% today.  

However, for enterprises to successfully integrate containers into their environment, they need a solution to tackle the complexity of deploying, managing, and scaling containerized applications. Kubernetes is a popular orchestration platform that many businesses use to support their container deployments. That said, Kubernetes was not specifically designed with enterprise constraints and requirements in mind. As enterprises increasingly look to invest in containerized infrastructure, it becomes important to understand the types of solutions available to orchestrate Kubernetes.

In case you aren’t already familiar, let’s start with brass tacks. A container is a standard unit of software that packages code and dependencies together. This way, an application will run more quickly and reliably, no matter the computing environment it encounters. Container images (which then become containers at runtime) are lightweight, standalone, executable packages of software with everything needed to run an application built right in, including:

• Code

• Runtime

• System tools

• System libraries 

• Settings

Containers are, by definition, self-contained. 

For additional background on containers, read our post: What is a container? Definition, benefits and use cases.

The challenge of containers is that they must be given commands for deployment, scaling, and management. Kubernetes arose as an open source container orchestration tool that automates all of these functions. Google developed Kubernetes internally, but they donated it to the Cloud Native Computing Foundation (CNCF), who now maintains it, in July 2015. 

Kubernetes single-handedly made containers accessible for a much wider variety of users. However, Kubernetes does not inherently put enterprise issues like reliability, ease of use, and productivity at the heart of its service. Therefore many enterprise Kubernetes users have turned to container management platforms to take their container use to the next level. Let’s take a look at what this means. 

There are two key gaps between containers and enterprise standards. These are most critical to larger enterprise customers, but many are important for much smaller businesses as well. Let’s take a look at each in detail.

Enterprises need to ensure they are not wasting computing instances when they move to containers. Without proper precautions it’s easy to utilize containers inefficiently, costing the business time and money while increasing complexity.

Many enterprises need to be able to move applications between different cloud platforms and dependencies. They have a responsibility to manage risk, and this means they must also ensure portability. They must be able to carry over all networking, security, and compliance policies and settings across application execution environments, which is not possible with Kubernetes out of the box. 

Kubernetes is highly opinionated, meaning that it is designed to work in a singular and specific way. While this may seem like it would make Kubernetes simpler to use, most enterprises have unique environments that require numerous adjustments and manual configurations during installation and set up. Simply put, Kubernetes on its own often isn’t enough for larger companies. 

The fact that Kubernetes is open source also means it evolves rather quickly. SDX Central states that there were 1.1 million total contributions to Kubernetes between its CNCF donation (July 2015) and August 2019. This rapid pace of change creates a good deal of inherent complexity in the many versions that can be chosen from at any time. Most enterprises require a more comprehensive on-ramp than Kubernetes comes with out of the box. 

The right Kubernetes orchestration solution enables efficiency through high application density (co-location) while also allowing flexible workload movement with minimal effort. It ensures that Kubernetes is properly and uniquely configured for each enterprise installation and allows users to easily port features across cloud providers, development teams, lines of business, or accounts. This portability protects the enterprise by allowing them to enforce their standards across all of their environments at once. 

When set up properly, Kubernetes is efficient, resilient, and nimble. However, enterprises come to the table with a unique and complex set of requirements. The best way to take full advantage of the many benefits Kubernetes has to offer is investing in an orchestration platform to alleviate  the pain caused by the gaps discussed above. 

Now that you’re thinking about container orchestration options, consider our very own! Critical Stack from Capital One allows users to get up and running with Kubernetes faster and reap more of its benefits. The developer tier is available to check out now, and our Enterprise Tier is slated for release later this year. If you’re new to containers, trying to solve resource gaps, or have a need for speed and automatic scaling, Critical Stack may be the right solution for your use case.