Frequently asked questions
Troubleshooting
Troubleshooting advice can be found on our Troubleshooting Cheatsheet.Configuration
How to Enable KubeSpan
Talos Linux provides a full mesh WireGuard network for your cluster.
To enable this functionality, you need to configure KubeSpan and Cluster Discovery in your Talos Linux configuration:
machine:
network:
kubespan:
enabled: true
cluster:
discovery:
enabled: false
Since KubeSpan encapsulates traffic into a WireGuard tunnel, Kube-OVN should also be configured with a lower MTU value.
To achieve this, add the following to the Cozystack ConfigMap:
apiVersion: v1
kind: ConfigMap
metadata:
name: cozystack
namespace: cozy-system
data:
values-kubeovn: |
kube-ovn:
mtu: 1222
Operations
How to enable access to dashboard via ingress-controller
Update your ingress application and enable dashboard: true option in it.
Dashboard will become available under: https://dashboard.<your_domain>
What if my cloud provider does not support MetalLB
You still have the opportunity to expose the main ingress controller using the external IPs method.
Take IP addresses of the external network interfaces for your nodes.
Add them to the externalIPs list in the Ingress configuration:
kubectl patch -n tenant-root ingresses.apps.cozystack.io ingress --type=merge -p '{"spec":{
"externalIPs": [
"192.168.100.11",
"192.168.100.12",
"192.168.100.13"
]
}}'
After that, your Ingress will be available on the specified IPs:
# kubectl get svc -n tenant-root root-ingress-controller
root-ingress-controller ClusterIP 10.96.91.83 37.27.60.28,65.21.65.173,135.181.169.168 80/TCP,443/TCP 133d
How to cleanup etcd state
Sometimes you might want to flush the etcd state from a node. You can use the following command:
talosctl reset --system-labels-to-wipe=EPHEMERAL --graceful=false --reboot
How to generate kubeconfig for tenant users
Use the following script:
user=tenant-root
cluster=$(kubectl config get-contexts | awk '$1 == "*" {print $3}')
token=$(kubectl get secret -n "$user" "$user" -o go-template='{{ printf "%s\n" (index .data "token" | base64decode) }}')
kubectl config view --minify --raw > tenant-kubeconfig
kubectl config --kubeconfig tenant-kubeconfig unset users
kubectl config --kubeconfig tenant-kubeconfig unset contexts
kubectl config --kubeconfig tenant-kubeconfig set "users.$user.token" "$token" --set-raw-bytes=true
kubectl config --kubeconfig tenant-kubeconfig set "contexts.$user@$cluster.user" "$user"
kubectl config --kubeconfig tenant-kubeconfig set "contexts.$user@$cluster.namespace" "$user"
kubectl config --kubeconfig tenant-kubeconfig set "contexts.$user@$cluster.cluster" "$cluster"
kubectl config --kubeconfig tenant-kubeconfig set current-context "$user@$cluster"
in the result, you’ll receive the tenant-kubeconfig file, which you can provide to the user.
How to configure Cozystack using FluxCD or ArgoCD
Here you can find reference repository to learn how to configure Cozystack services using GitOps approach:
How to rotate CA
In general, you almost never need to rotate the root CA certificate and key for the Talos API and Kubernetes API. Talos sets up root certificate authorities with the lifetime of 10 years, and all Talos and Kubernetes API certificates are issued by these root CAs. So the rotation of the root CA is only needed if:
- you suspect that the private key has been compromised;
- you want to revoke access to the cluster for a leaked talosconfig or kubeconfig;
- once in 10 years.
For tenant k8s cluster:
See: https://kamaji.clastix.io/guides/certs-lifecycle/
export NAME=k8s-cluster-name
kubectl delete secret ${NAME}-ca
kubectl delete secret ${NAME}-sa-certificate
kubectl delete secret ${NAME}-api-server-certificate
kubectl delete secret ${NAME}-api-server-kubelet-client-certificate
kubectl delete secret ${NAME}-datastore-certificate
kubectl delete secret ${NAME}-front-proxy-client-certificate
kubectl delete secret ${NAME}-konnectivity-certificate
kubectl delete secret ${NAME}-admin-kubeconfig
kubectl delete secret ${NAME}-controller-manager-kubeconfig
kubectl delete secret ${NAME}-konnectivity-kubeconfig
kubectl delete secret ${NAME}-scheduler-kubeconfig
kubectl delete po -l app.kubernetes.io/name=kamaji -n cozy-kamaji
Wait for virt-launcher-kubernetes-* pods restart. Download new k8s certificate.
For managment k8s cluster:
See: https://www.talos.dev/v1.9/advanced/ca-rotation/#kubernetes-api
git clone https://github.com/cozystack/cozystack.git
cd packages/core/testing
make apply
make exec
Add to your talosconfig in pod:
client-aenix-new:
endpoints:
- 12.34.56.77
- 12.34.56.78
- 12.34.56.79
nodes:
- 12.34.56.77
- 12.34.56.78
- 12.34.56.79
Exec in pod:
talosctl rotate-ca -e 12.34.56.77,12.34.56.78,12.34.56.79 --control-plane-nodes 12.34.56.77,12.34.56.78,12.34.56.79 --talos=false --dry-run=false &
Get new kubeconfig:
talm kubeconfig kubeconfig -f nodes/srv1.yaml
For talos API
See: https://www.talos.dev/v1.9/advanced/ca-rotation/#talos-api All like for managment k8s cluster, but talosctl command:
talosctl rotate-ca -e 12.34.56.77,12.34.56.78,12.34.56.79 --control-plane-nodes 12.34.56.77,12.34.56.78,12.34.56.79 --kubernetes=false --dry-run=false &
Bundles
How to overwrite parameters for specific components
Moved to the Bundles reference.
How to disable some components from bundle
Moved to the Bundles reference.