AWS self-managed k8s cluster directly on ec2 by ansible
Rancher Support matrix
Local docker Installation
To enable Rancher on homelab env, we need a Linux box to run Rancher as docker.
Rancher Web Portal login
via https://localhost/dashboard/home
Import existing k8s culster vs create new k8s from rancher console
Under “cluster management”, it supports importing k8s from cloud providers to local k8s, unfutinately I previous k8s cluster is v1.28 which is too high to be imported and managed by this rancher.
Hence I will use Rancher to create a new one here First prepare 3 local Linux VM boxes, come back to Rancher console under cluster management, give name of the new cluster, then run command to initiate control plane.
Updating new machine as a K8S rancher node as control plane.
Then join the 2 worker nodes
Create zackweb and joesite as deployment from Rancher console
Conclusion
now we can use Rancher to deploy a local k8s cluster based on 3 Linux machines without any trouble just a few commands. then we will be able to create deployment and service in rancher console instead of “kubectl” all the time, it also provides app market for most popular helm charts ready to be installed just by one click like Istio and Prometheus. Only downside is, Rancher itself requires resources to run which may impact the performance and resources on each node, also it brings complexity in upgrade for both Rancher and k8s. overall I love the concept and tools that Rancher provides to manage k8s cluster. I will explore more in the next blog.