Steps for migrating a single-node 2.1.x installation to HA

migrate-rancher-2-1-x-single-node

docker run -d --restart unless-stopped --name rancher -p 80:80 -p 443:443 -v /host/rancher:/var/lib/rancher rancher/rancher:v2.1.5

docker exec -it rancher /bin/bash

# cd /var/lib/rancher
# tar -zcvf pki.bundle.tar.gz -C /etc/kubernetes/ssl .
# cp -r /etc/kubernetes/ssl .
# exit

docker run --net=container:$(docker ps | grep rancher | awk '{print $1}') -it --volumes-from rancher rancher/rke-tools:v0.1.20

# ETCDCTL_API=3 etcdctl --endpoints=127.0.0.1:2379 --cacert=/var/lib/rancher/ssl/kube-ca.pem --cert=/var/lib/rancher/ssl/kube-etcd-127-0-0-1.pem --key=/var/lib/rancher/ssl/kube-etcd-127-0-0-1-key.pem snapshot save /var/lib/rancher/single-node-snapshot
# exit

exit

scp -r root@192.168.100.135:/host/rancher/pki.bundle.tar.gz .
scp -r root@192.168.100.135:/host/rancher/single-node-snapshot .

ssh root@192.168.100.137 # new node
mkdir -p /opt/rke/etcd-snapshots

scp single-node-snapshot root@192.168.100.137:/opt/rke/etcd-snapshots/
scp pki.bundle.tar.gz root@192.168.100.137:/opt/rke/etcd-snapshots/

### You must create a new cluster.yml with your new desired node in it

rke etcd snapshot-restore --name single-node-snapshot

rke up

export KUBECONFIG=$(pwd)/kube_config_cluster.yml

kubectl -n kube-system create serviceaccount tiller

kubectl create clusterrolebinding tiller \
  --clusterrole cluster-admin \
  --serviceaccount=kube-system:tiller

helm init --service-account tiller

helm repo add rancher-stable https://releases.rancher.com/server-charts/stable

helm repo update

helm install stable/cert-manager --name cert-manager --namespace kube-system

helm install rancher-stable/rancher --name rancher --namespace cattle-system --set hostname=ck-sn2ha.fe.rancher.space