Hi Team, I'm new to the group. Just have few clarifications.
1. when we deploy Rancher Management in HA, we generally deploy three different VM and connect them to L4 Load Balancer with VIP. Eventually all the three VM's will be running with the server rke2-server. If thats the case, Do we need to rke2-agent deployment on these machines , If not, what is the use case of deploying rke2-agent?
02/03/2023, 11:53 AM
Running 3 rke2-server instances means you are running the control plane in HA, which is correct. The rke2-agent instances would be for "worker only" nodes.
02/03/2023, 3:30 PM
@ancient-army-24563 Going tangential here. Which L4 loadbalancer did you use and what ports are they forwarding TCP traffic to rke2?
02/03/2023, 6:03 PM
@big-hydrogen-97240 Yes great, Thanks for your response. makes sense
@stale-spring-20280 im deploying in the cloud provided called G42 in MiddleEast. So im using their Elastic Load Balance (L4) and the port they’re listening is 80 and backend servers (3 VMs) with 443
also we might need to 6443 and change the kubeconfig cluster ip to VIP ip address so that we connect with cluster ip to kube-apiserver