Consul on k8s via Helm: Client and server pods cannot be scheduled on the same node

24 Views Asked by At

According to consul, if choosing to expose the k8s consul servers to the LAN so LAN vm's can register with k8s Consul: this needs to be done

server:
  enabled: true
  exposeGossipAndRPCPorts: true
  ports:
    serflan:
      port: 9301

The switch to port 9301 is done so not colide with agent pod that's running on the same node as the server pod.

Which makes sense: as we're moving the server serflan to a different port, to avoid collision with the client pod's serflan port.

When doing so however, only one of the two pods gets actually scheduled, while the other:

0/1 nodes are available: 1 node(s) didn't have free ports for the requested pod ports

If anyone has succeeded with running server and client Consul pods on the same node - or another solution that enables having a consul daemonset on each k8s node, please share your idea, thank you.

0

There are 0 best solutions below