Hello all,
i’m a newbie for Kubernetes and trying to understand kubernetes concepts.
please correct if i get this concept wrong, “replicaSet offer self-healing for PODs and pods get recreated if any one die, deleted or not accessible.”
I have a k8 cluster setup in home lab with one master and 3 worker nodes where I deployed one replicaset with 10 replicas and it was distributed across all the worker nodes. Later I shutdown one of the worker node and wanted to see the impact. I was expecting it should recreate those pods on running worker nodes which was now not accessible due to one of worker node down.
I waited almost 5 minutes then new pods was created on the remaining running worker node to maintain desire state of RS.
My question is why it is taking so long to do so?
wondering if it is expected behavior.
Cluster information:
Kubernetes version: 1:26
Cloud being used: bare-metal
Installation method: Manual Home-lab
Host OS: Ubuntu 22.04 LTE desktop
CNI and version: 10-weave.conflist
CRI and version: containerd.io 1.6.21
Wondering if there is any limitation for self-healing properties.