Cluster information:
Kubernetes version: k3s version 1.19.1
Cloud being used: (put bare-metal if not on a public cloud) : On contabo VPS
Installation method: GitHub - techno-tim/k3s-ansible: The easiest way to bootstrap a self-hosted High Availability Kubernetes cluster. A fully automated HA k3s etcd install with kube-vip, MetalLB, and more. Build. Destroy. Repeat.
Host OS: Ubuntu version 22.04
CNI and version: flannel
MORE DETAILS HERE : How to contact a VIP kube-vip in a HA cluster under VPN wireguard · Issue #617 · techno-tim/k3s-ansible · GitHub
Hello,
I have 5 VPS I wanted to have some fun and make a kubernetes HA cluster. I am a beginner in the world of kubernetes.
So my nodes communicate through a vpn tunnel under wireguard.
I have 3 masters 10.0.0.1 and 10.0.0.2 and 10.0.0.3
and 2 workers in 10.0.0.3 and 10.0.0.5.
and a VIP (kube-vip) in 10.0.0.200
all my flows go through the wg0 interface for traffic in 10.0.0.0/24
The problem is that my worker and master nodes manage to communicate with each other via the VPN, but when I decide to have my workers communicate with the VIP, there’s no response from the VIP.
I think I’m misconfiguring the Kube-VIP in my cluster.
I’m also wondering about using BGP to have dynamic routes depending on the nodes and for HA.
If someone can explain me the BGP with Kube-vip or how can i solve the problem please
The ansible remains blocked when k3s-nodes has to fetch a curl of the 10.0.0.200 cert:
k3s-node service
~# kubectl get pods -n kube-system
NAME READY STATUS RESTARTS AGE
coredns-576bfc4dc7-8hdm2 1/1 Running 1 (8m48s ago) 10m
kube-vip-ds-bppf4 1/1 Running 1 (8m41s ago) 9m28s
kube-vip-ds-r7r8t 1/1 Running 1 (8m48s ago) 10m
kube-vip-ds-wnmp8 1/1 Running 1 (8m49s ago) 9m55s
local-path-provisioner-86f46b7bf7-gw2zr 1/1 Running 1 (8m48s ago) 10m
metrics-server-557ff575fb-tn9d5 1/1 Running 1 (8m48s ago) 10m