Open spisarski opened 6 years ago
Added the initial check of the node count. Now need to expand further with the following information suggested by @sbhojpur via Slack
Sachin Bhojpuria [3:12 AM] @spisarski: I think we should check for the following to validate our k8s installation Components Pods Containers Node Tolerance Status
kube-apiserver-xxxxxx 1 1 Master No restart Running
kube-controller-manager-xxxxxx 1 1 Master No restart Running
kube-dns-xxxxx 1 3 Master No restart Running
kube-dns-xxxxx 1 3 Minion No restart Running
Kube-proxy-xxxxx 1 1 Master No restart Running
Kube-proxy-xxxxx 1 1 All Minions No restart Running
kube-scheduler-xxxxx 1 1 Master No restart Running
kubedns-autoscaler-xxxxx 1 1 Master No restart Running
nginx-proxy-xxxxxx 1 1 All Minions No restart Running
etcd.service Run as linux daemon services on master node
Kubelet.service Run as linux daemon services on each node
Networking
Weave-net-xxxxx 1 2 Master No restart Running
Weave-net-xxxxx 1 2 All Minions No restart Running
Flannel-xxxx 1 2 Master No restart Running
Flannel-xxxx 1 2 All Minions No restart Running
Kubectl get networks Get list of all networks created by SNAPS-kubernetes, it should match the networks defined under Multus
Storage
Kubectl get PVC It should list all claims defined in the yaml and all claims should be in bound state
Kubectl get PV It should list volumes defined, and they should be in bound state
More validation has been added but not enough and still need to implement cleanup as well
Test the iaas_launch.py -k8_c option and ensure 0 return to OS and verify all expected services and packages are removed.