CLI commands¶

Run infrastructure commands as root on the connected server. Infrastructure 7 configures /root/.kube/config during installation. If the default config is unavailable, add --kubeconfig=/etc/kubernetes/admin.conf to each kubectl command.

Infrastructure 7¶

Check cluster status¶

kubectl get nodes -o wide
kubectl get pods --all-namespaces -o wide

Wodby's system workloads run in the wodby namespace:

kubectl -n wodby get pods -o wide

Restart Wodby Agent¶

kubectl -n wodby rollout restart deployment/agent
kubectl -n wodby rollout status deployment/agent --timeout=5m

Updating Infrastructure 7 Agent¶

Use the Dashboard infrastructure update when it is offered for your server. Infrastructure 7.0.2 uses Agent 5.6.0. Dashboard application-log polling requires Agent 5.4.2 or newer.

Run the following manual update only when instructed by Wodby support, with no Dashboard infrastructure update in progress. Server management briefly disconnects while Agent restarts; application containers and Edge continue running.

kubectl -n wodby set image deployment/agent agent=wodbycloud/agent:5.6.0
kubectl -n wodby rollout status deployment/agent --timeout=5m

Verify the configured image:

kubectl -n wodby get deployment agent -o jsonpath='{.spec.template.spec.containers[0].image}{"\n"}'

Confirm that the server reconnects in the Dashboard. Changing the image manually does not advance the recorded infrastructure version or complete a pending Dashboard update. If the server remains disconnected or in maintenance, contact Wodby support.

Restart Edge¶

Edge owns the server's public ports, so restarting it can cause a brief interruption:

kubectl -n wodby rollout restart deployment/edge
kubectl -n wodby rollout status deployment/edge --timeout=5m

View system workload logs¶

kubectl -n wodby logs deployment/agent --since=30m --timestamps
kubectl -n wodby logs deployment/edge --since=30m --timestamps
kubectl -n wodby logs statefulset/etcd --since=30m --timestamps

Restart host runtime services¶

Restarting these services interrupts applications on the server. Take a provider snapshot and verify that no maintenance operation is running first.

systemctl restart docker
systemctl restart cri-docker.socket kubelet
kubectl get nodes
kubectl get pods --all-namespaces

Host service logs¶

journalctl -u kubelet -u docker -u cri-docker.service --since "30 minutes ago"
journalctl -u wodby-installer --since "30 minutes ago"

Check installed component versions¶

cat /opt/wodby/etc/infrastructure-7.env
kubectl version
docker version
kubectl -n wodby get deployment agent edge -o custom-columns=NAME:.metadata.name,IMAGE:.spec.template.spec.containers[0].image

The Kubernetes deployment image is authoritative for Agent and Edge patch versions on a running server; the recorded installation profile describes what the installer originally applied.

Clean up failed task Pods¶

Review the list before deleting anything:

kubectl get pods --all-namespaces | grep 'task-' | grep -E 'Error|Failed'

Delete a confirmed failed task Pod with:

kubectl -n INSTANCE_UUID delete pod POD_NAME

Infrastructure 6 and older¶

The following commands are retained for existing legacy servers. Do not use the Infrastructure 7 systemd or kubeadm assumptions on these hosts.

Restart Wodby Agent¶

docker rm -f $(docker ps | grep wodbycloud/agent | awk '{ print $1 }')

Restart Edge¶

docker rm -f $(docker ps | grep wodby/edge-alpine | awk '{ print $1 }')

Restart Docker and Kubernetes services¶

systemctl stop kube-apiserver
systemctl stop kube-controller
systemctl stop kube-kubelet
systemctl stop kube-proxy
systemctl stop kube-scheduler
systemctl stop docker

systemctl start docker
systemctl start kube-apiserver
systemctl start kube-controller
systemctl start kube-kubelet
systemctl start kube-proxy
systemctl start kube-scheduler

Check node and namespaces¶

kubectl describe node 127.0.0.1
kubectl get namespaces

General host checks¶

df -h
vmstat -Sm 1
journalctl -f