Summary
We started this chapter by installing Argo CD with Kustomize. We went with a cloud provider-managed cluster because we needed more nodes to experiment with an HA deployment. We experienced how Argo CD can update itself and how we can make configuration changes to our installation. While a production Kubernetes cluster is highly available and a cloud provider will manage it for us, there are still scenarios when disasters can happen, so we need to have a working disaster recovery strategy. We saw how we can create backups of the Argo CD state and then restore it in a new cluster.
Observability is an important topic and we discussed which metrics can be used to monitor an Argo CD installation, from the OOM container restarts to what a microservice team needs to watch out for. We finished by learning how to link the result of a synchronization to a pipeline so that everything will be automated.
In the next chapter, we are going to discover how we can bootstrap a new Kubernetes...