Packaging Cloud Native Apps For Kubernetes With Helm

Key Takeaways

To deploy applications on Kubernetes, we can use YAML manifests. These manifests are relatively simple, straightforward to deploy, but they lack in flexibility. How can we adapt them to our environment, without doing manual editing or maintaining multiple versions side by side? How can we keep track of successive deployed versions, and roll back if a deployment doesn't go well? How can we distribute (and find!) these manifests and their successive revisions?

Helm addresses a lot of these challenges. Our Kubernetes manifests can be extended with a powerful templating language and packaged as Helm "charts". These charts can expose a number of tunable parameters (or "values"), distributed through repositories, indexed by a software Hub. Helm automatically keeps track of the successive releases that we deploy, allowing us to roll back if necessary. Furthermore, Helm is a "graduate project" in the Cloud Native Computing Foundation, meaning that it's now a stable and mature project.

In this hands-on workshop, we will learn how to use Helm to install and configure third-party apps, and then we will learn to write our own Helm charts. We will cover Helm concepts, its templating language, and dive into a few examples of increasing complexity to see some best practices as well as useful tips and tricks.

SPEAKER