Confirming the deployment#
Note
After running the helm install command, please wait 5 to 10 minutes for the deployment to stabilize.
To understand whether installation is complete, start by making sure all pods are ready in the kcm-system namespace. There should be 20 pod entries:
kubectl get pods -n kcm-system
NAME READY STATUS RESTARTS AGE
azureserviceoperator-controller-manager-6b4dd86894-4dpfc 1/1 Running 0 7m15s
capa-controller-manager-64bbcb9f8-ltj6z 1/1 Running 0 6m58s
capd-controller-manager-7586b6577c-6w4wq 1/1 Running 0 7m11s
capg-controller-manager-774958b9b9-hvxvv 1/1 Running 0 6m54s
capi-controller-manager-5b67d4fc7-dhk7p 1/1 Running 0 8m49s
capo-controller-manager-6f98bb68cd-vsc24 1/1 Running 0 6m33s
capv-controller-manager-69f7fc65d8-jx8s4 1/1 Running 0 6m40s
capz-controller-manager-5b87fdf745-mzm7x 1/1 Running 0 7m15s
helm-controller-746d7db585-x64ld 1/1 Running 0 11m
k0smotron-controller-manager-bootstrap-67dd88d848-wfrhm 2/2 Running 0 8m32s
k0smotron-controller-manager-control-plane-657f5578d4-7vb8t 2/2 Running 0 8m24s
k0smotron-controller-manager-infrastructure-5867d575f9-t28hp 2/2 Running 0 6m46s
kcm-cert-manager-6979c67bc4-b6s4w 1/1 Running 0 11m
kcm-cert-manager-cainjector-5b97c84fdb-kdsw5 1/1 Running 0 11m
kcm-cert-manager-webhook-755796f599-q6727 1/1 Running 0 11m
kcm-cluster-api-operator-65c8f75569-rfsp2 1/1 Running 0 9m30s
kcm-controller-manager-68b56bff85-6fmsp 1/1 Running 0 9m30s
kcm-velero-67bf545995-x6784 1/1 Running 0 11m
source-controller-74b597b995-kkqqw 1/1 Running 0 11m
kubectl get pods -n kcm-system --no-headers | wc -l
20
State management is handled by Project Sveltos, so you'll want to make sure that all 10 pods are running/completed in the projectsveltos namespace:
kubectl get pods -n projectsveltos
NAME READY STATUS RESTARTS AGE
access-manager-6696df779-pnxjx 1/1 Running 0 10m
addon-controller-6cb6c5f6df-zmfch 1/1 Running 0 10m
classifier-manager-5b47b66fc9-5mtwl 1/1 Running 0 10m
event-manager-564d6644b4-wr9cq 1/1 Running 0 10m
hc-manager-7c56c59d9c-w5gds 1/1 Running 0 10m
sc-manager-6798cd9d4d-r7z9j 1/1 Running 0 10m
shard-controller-797965bb58-65lmp 1/1 Running 0 10m
sveltos-agent-manager-5445f6f57c-wxw2s 1/1 Running 0 10m
techsupport-controller-5b666d6884-jfqnp 1/1 Running 0 10m
kubectl get pods -n projectsveltos --no-headers | wc -l
10
If any of these pods are missing, simply give k0rdent more time. If there's a problem, you'll see pods crashing and restarting, and you can see what's happening by describing the pod, as in:
kubectl describe pod classifieclassifier-manager-5b47b66fc9-5mtwl -n projectsveltos
As long as you're not seeing pod restarts, you just need to wait a few minutes.
Verify that k0rdent itself is ready#
The actual measure of whether k0rdent is ready is the state of the Management object. To check, issue this command:
kubectl get Management -n kcm-system
NAME READY RELEASE AGE
kcm True kcm-0-3-0 9m
Verify the templates#
Next verify whether the KCM templates have been successfully installed and reconciled. Start with the ProviderTemplate objects:
kubectl get providertemplate -n kcm-system
NAME VALID
cluster-api-0-3-0 true
cluster-api-provider-aws-0-3-0 true
cluster-api-provider-azure-0-3-0 true
cluster-api-provider-docker-0-3-0 true
cluster-api-provider-gcp-0-3-0 true
cluster-api-provider-k0sproject-k0smotron-0-3-0 true
cluster-api-provider-openstack-0-3-0 true
cluster-api-provider-vsphere-0-3-0 true
kcm-0-3-0 true
projectsveltos-0-50-0 true
Make sure that all templates are not just installed, but valid. Again, this may take a few minutes.
You'll also want to make sure the ClusterTemplate objects are installed and valid:
kubectl get clustertemplate -n kcm-system
NAME VALID
adopted-cluster-0-3-0 true
aws-eks-0-3-0 true
aws-hosted-cp-0-3-0 true
aws-standalone-cp-0-3-0 true
azure-aks-0-3-0 true
azure-hosted-cp-0-3-0 true
azure-standalone-cp-0-3-0 true
docker-hosted-cp-0-3-0 true
gcp-gke-0-3-0 true
gcp-hosted-cp-0-3-0 true
gcp-standalone-cp-0-3-0 true
openstack-standalone-cp-0-3-0 true
remote-cluster-0-3-0 true
vsphere-hosted-cp-0-3-0 true
vsphere-standalone-cp-0-3-0 true
If this is a new v0.2.0 cluster, you will see:
No resources found in kcm-system namespace.
You can then install the ServiceTemplates objects as follows:
helm upgrade --install catalog-core oci://ghcr.io/k0rdent/catalog/charts/catalog-core --version 1.0.0 -n kcm-system
Finally, make sure the ServiceTemplate objects are installed and valid:
kubectl get servicetemplate -n kcm-system
NAME VALID
cert-manager-1-16-2 true
dex-0-19-1 true
external-secrets-0-11-0 true
ingress-nginx-4-11-0 true
ingress-nginx-4-11-3 true
kyverno-3-2-6 true
velero-8-1-0 true