I’ve developed an ansible playbook that can deploy a cluster into a pre-existing RKE cluster, however, the second master ends up in an invalid state.
This is all being done thru the rancher cli.
In a nutshell it does:
- create cluster
- compiles master, etcd and worker join commands
- runs the join commands on each respective node
This works fine except the secondary master issue. Deployed a 5-node etcd, 2-node master, and 50-node worker cluster.
The secondary master is the last piece I need to iron out.
Can anyone point me in the right direction?