profile
viewpoint

Ask questionsrke 0.1.17 Can't initiate NewClient: protocol not available

I updated from rke 0.1.15 and tried to update our rancher installation to kubernetes 1.11.8 and received the following error. Downgrading from 0.1.17 back to 0.1.15 allows rke to connect again.

RKE version: rke 0.1.17 on windows 10 professional

Steps to Reproduce:

$ rke up --config ~/rancher.yml

Results:

time="2019-03-08T20:31:04-08:00" level=info msg="Building Kubernetes cluster"
time="2019-03-08T20:31:04-08:00" level=info msg="[dialer] Setup tunnel for host [172.x.x.x]"
time="2019-03-08T20:31:04-08:00" level=warning msg="Failed to set up SSH tunneling for host [172.x.x.x]: Can't initiate NewClient: protocol not available"
time="2019-03-08T20:31:04-08:00" level=info msg="[dialer] Setup tunnel for host [172.x.x.x]"
time="2019-03-08T20:31:04-08:00" level=warning msg="Failed to set up SSH tunneling for host [172.x.x.x]: Can't initiate NewClient: protocol not available"
time="2019-03-08T20:31:04-08:00" level=info msg="[dialer] Setup tunnel for host [172.x.x.x]"
time="2019-03-08T20:31:04-08:00" level=warning msg="Failed to set up SSH tunneling for host [172.x.x.x]: Can't initiate NewClient: protocol not available"
time="2019-03-08T20:31:04-08:00" level=warning msg="Removing host [172.x.x.x] from node lists"
time="2019-03-08T20:31:04-08:00" level=warning msg="Removing host [172.x.x.x] from node lists"
time="2019-03-08T20:31:04-08:00" level=warning msg="Removing host [172.x.x.x] from node lists"
time="2019-03-08T20:31:04-08:00" level=fatal msg="Cluster must have at least one etcd plane host: failed to connect to the following etcd host(s) [172.x.x.x]"
rancher/rke

Answer questions robkaandorp

Work with ubuntu on windows 10 https://docs.microsoft.com/en-us/windows/wsl/install-win10

I can verify, this indeed works with version 0.1.18, but you will have to use the linux version of rke.

Start an Ubuntu shell in windows, go to the correct drive and path (you will find your C: drive under /mnt/c) and run ./rke_linux_amd64 up.

useful!

Related questions

"Failed to reconcile etcd plane" when updating RKE binary hot 3
Failed to get /health for host - remote error: tls: bad certificate hot 2
Failed to rotate expired certificates on an RKE cluster: unable to reach api server to fetch CA hot 2
Error response from daemon: chown /etc/resolv.conf: operation not permitted hot 1
Pods can't reach networks outside of node hot 1
Calico node failed to start after upgrading the cluster hot 1
Job rke-network-plugin-deploy-job never completes (virtualbox) hot 1
rke up --local fails to deploy successfully hot 1
Job rke-network-plugin-deploy-job never completes (virtualbox) hot 1
go panic on intial rke up hot 1
Unable to update cluster "crypto/rsa: verification error" hot 1
Calico node failed to start after upgrading the cluster hot 1
pods in "CrashLoopBackOff" status after restoring from backup hot 1
[SOLVED] Failed to apply the ServiceAccount needed for job execution: Post https://10.102.X.X:6443/apis/rbac.authorization.k8s.io/v1/clusterrolebindings: Forbidden hot 1
Failed to get /health for host - remote error: tls: bad certificate hot 1
Github User Rank List