Test Cluster Configuration

The Development cluster is deployed using K3s - Kubernetes Simplified.

These packages form the basic functionality of the development cluster. The packages need to preserve the IP address assignments. It is expected that the Test Cluster will be destroyed and re-created as needed to test configurations and processes/procedures.

Scripts & config files are checked into gitlab under the Kubernetes group project listed.

activity	gitlab	script/procedures/config	IP	hostname(s)
K8Dash Dashboard	k8s-admin		10.0.0.200
NVIDIA device plugin		https://github.com/NVIDIA/k8s-device-plugin

Dashboard Token

Obtain the current dashboard token with this command:

kubectl describe secrets \
  `kubectl get secrets | awk '/k8dash-sa/ {print $1}'` \
  | awk '/token:/ {print $2}'

The current Test cluster dashboard token is:

eyJhbGciOiJSUzI1NiIsImtpZCI6IktfVTJCLU9xVG1zRmVOTWF5dTBZcWNSWlVzQ2t0RlFaS0o4VEFmcUl2VE0ifQ.eyJpc3MiOiJrdWJlcm5ldGVzL3NlcnZpY2VhY2NvdW50Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9uYW1lc3BhY2UiOiJkZWZhdWx0Iiwia3ViZXJuZXRlcy5pby9zZXJ2aWNlYWNjb3VudC9zZWNyZXQubmFtZSI6Ims4ZGFzaC1zYS10b2tlbi16czJkNCIsImt1YmVybmV0ZXMuaW8vc2VydmljZWFjY291bnQvc2VydmljZS1hY2NvdW50Lm5hbWUiOiJrOGRhc2gtc2EiLCJrdWJlcm5ldGVzLmlvL3NlcnZpY2VhY2NvdW50L3NlcnZpY2UtYWNjb3VudC51aWQiOiIzYzg5OGUyMC0wNGFmLTQ5YTItYWJmYS01MTcwOGE5OGYxMTUiLCJzdWIiOiJzeXN0ZW06c2VydmljZWFjY291bnQ6ZGVmYXVsdDprOGRhc2gtc2EifQ.dkxXHdHFrwJ2krjd3krFHFM5aAMLIRvDH8cFv9TzaCRgolnlC-xI12rfLXnyig-WKWdjzJSUr-_bh60CRmzONBzXWg9NEY_LYJRIKszkGOGADFPNuCn-ugbfq5LqvEh5H2LWsnzWx49YUQeHNeVjipRKl9xLSqvmK2UfmvffJYFoxcnuky4xC-X4qv8aYzXSF9D9hPR7NYd8XdFv3NZsPOOkhwjBpKZi20FEKLeuNNJq88Xdewh1mGLkCyeZuSFpNV_eawOJSO4A2NpT5hCmzleq4XDBNMZUlgALpTqhNd5kGJnke6xt8WIt2EP3HarRkq1wLLu5TIR5oB2xQ2qR8w

Storage

The test cluster does not (by default) have any required filesystems, but systems that require access to both the development filesystem (/workspace) and the production filesystem (/shared) can mount them via ceph or NFS as appropriate.

There is a test ceph cluster hosted on pro5 that is available through the rook installation, but (at this time) no filesystems are hosted on that cluster.

Kubernetes Node Join Command

kubeadm join 10.0.0.70:6443 --token w0vbly.4ze7o9rthhinav6d \
    --discovery-token-ca-cert-hash sha256:cc6870fc970691dc2134cef21f15fa55ca001c8edaf98df11c9fe84457c29dbd \
    --ignore-preflight-errors Swap --node-name=`hostname -s`