After sharing some context details about how I manage my personal k3s cluster. I would share a different way to create helm charts that heavily relies on a library chart and makes everything configurable via helm values. This approach is the one that the k8s-at-home project uses (https://docs.k8s-at-home.com/our-helm-charts/common-library/).
Recently, we migrated the Atlas Rucio installation from puppet to Kubernetes. This talk will presents how we use terraform, flux2 and helm to manage our multi-cluster installation and the applications inside the clusters.
CERN's Web Frameworks are supported by a set of OpenShift clusters with shared components. We'd like to share how we use GitOps to manage their configuration, relying on ArgoCD, Helm charts and Gitlab CI. We'd also like to discuss how we can improve our current approach by relying more on ArgoCD, and upcoming solutions for secrets management.
The multiple supported versions of the CERN...
In this session you will learn how JEEDY team, build its internal infrastructure profiting of ArgoCD and git.
In particular we will show how we managed:
- monitoring infrastructure for both us and users
- submission of user cronjobs
The session will go through advantages, pitfalls of the GitOps approach that we faced during our development.
There are many, many options to handle secrets in Kubernetes deployments, and even more when considering handling sensitive data in GitOps setups.
In this short presentation we present the differences between handling (encrypted) data in Git, secrets inside the clusters, and the different tools available to simplify these tasks.
