I created kubernetes cluster on aws ec2 using kubeadm. Now I need to autoscale the K8's cluster when there are not enough resources on nodes to schedule new pods, How can I achieve autoscaling feature for my cluster?
Unfortunately there isn't a great answer if you mean you manually ran kubeadm
on some EC2 instance. cluster-autoscaler
is the thing to use, but it requires you deploy your nodes using Autoscaling Groups. It's possible to use ASGs and kubeadm
but I don't know of anything off-the-shelf for it.