eoli3n
commented
4 years ago Hi, When i add these routes on each workers, my pods can't reach outside world.
ip route add $worker1_pod_subnet via $worker1_ip
ip route add $worker2_pod_subnet via $worker2_ip
ip route add $worker3_pod_subnet via $worker3_ipI did some tests and found that the route to reach the current pod_subnet is conflicting with an existing one.
If you run a pod with crictl manually before adding routes on each, it will creates a default one as
$pod_subnet dev cnio0 proto kernel scope link src $pod_subnet_gateway More concretly i get this
root@tspeda-k8s-worker3 ~# ip route
default via 162.38.60.100 dev ens3 proto static
10.200.204.0/24 via 162.38.60.206 dev ens3
10.200.205.0/24 via 162.38.60.206 dev ens3
# The next one is automatically created
10.200.206.0/24 dev cnio0 proto kernel scope link src 10.200.206.1
162.38.60.0/24 dev ens3 proto kernel scope link src 162.38.60.206
# If you add this route, it will break cnio0 routing
# 10.200.206.0/24 via 162.38.60.206 dev ens3 So for each of you workers don't add the current pod subnet route.
V0idk
Hi there, I am following this guide on my virtual machine cluster, and I got stuck at 11-pod-network-routes.md, how can I create the routes so that a pod can ping another pod on a different node? I ran below 3 commands on each of the worker nodes, however it didn't work out.
route add -net 10.200.0.0 netmask 255.255.255.0 gw {worker-0 ip} route add -net 10.200.1.0 netmask 255.255.255.0 gw {worker-1 ip} route add -net 10.200.2.0 netmask 255.255.255.0 gw {worker-2 ip}
any suggestions?
info you might need: OS: Ubuntu 18.04.1 LTS kubernetes: v1.12.0
Thanks