Unable to deploy cluster using vSphere driver "Failed to apply the ServiceAccount"

Rancher 2.x
1

Hi,

I’m running a Rancher Server for testing purposes. I already provisioned a custom cluster on 3 custom nodes (all CentOS 7.7 1908 VMs, with Docker 18.09.9) that runs very fine. But trying to provision a Cluster with the vSphere driver is not working.

My environment

  • Single Node Rancher Test Environment
  • Running Rancher v2.3.2 on CentOS 7.7 1908 and Docker 18.09.9 installed on a VM
  • Our vSphere Environment is a vSphere 6.7
  • I’m using a recognized CA signed certificate
  • The whole system is running behind a Proxy
  • CloudCredentials for vSphere
  • Node Template with vSphere Credentials

This is the Rancher Template section from my Node Template:

69874061-5c4d2c80-12ba-11ea-9e45-b86a61b1c430
69874061-5c4d2c80-12ba-11ea-9e45-b86a61b1c4301087×918 46.6 KB

My cloud-config.yml file is hosted on a local webserver. But I also tried it without the cloud-config, the problem stays the same. The file looks like that:

#cloud-config
rancher:
  network:
    http_proxy: http://192.168.100.101:8080
    https_proxy: http://192.168.100.101:8080
    no_proxy: localhost,127.0.0.1,0.0.0.0,.our.domain,172.16.0.0/21

runcmd:
  - echo 172.16.1.245 ranchermgmt.our.domain >> /etc/hosts
    
ssh_authorized_keys:
  - ssh-rsa <my key> me@there

The deployment process runs without any errors until it comes to this point:

2019/11/29 13:28:45 [ERROR] cluster [c-nt6mr] provisioning: Failed to apply the ServiceAccount needed for job execution: Post https://172.16.3.19:6443/apis/rbac.authorization.k8s.io/v1/clusterrolebindings?timeout=30s: net/http: request canceled while waiting for connection (Client.Timeout exceeded while awaiting headers)
2019/11/29 13:28:46 [INFO] kontainerdriver rancherkubernetesengine stopped
2019/11/29 13:28:46 [ERROR] ClusterController c-nt6mr [cluster-provisioner-controller] failed with : Failed to apply the ServiceAccount needed for job execution: Post https://172.16.3.19:6443/apis/rbac.authorization.k8s.io/v1/clusterrolebindings?timeout=30s: net/http: request canceled while waiting for connection (Client

Afterwards rancher is trying to deploy the cluster again and again until I stop it.

I’m totally stuck with this. Thanks in advance for your help.