Cluster Cheatsheet

This page contains a useful selection of commands to troubleshoot and manage your cluster.

Fluxcd

Show all Flux Resources:

flux get all -A

Show all Kustomizations:

flux get ks -A

Show all Helm-Releases:

flux get hr -A

Reconcile your cluster:

flux reconcile source git cluster

Show all Flux objects which are not ready:

flux get all -A --status-selector ready=false

Helm

Get Values from a running deployment:

helm get values -n <namespace> <deployment-name>

Kubectl

Most of the following commands can be adapted to either show all namespaces -A or only show a specific namespace with -n <namespace>. Examples use the all namespaces flag. Additionally you can use additional commands like grep to sort and filter your results.

Show all Running Pods

kubectl get pods -A

Show all Services:

kubectl get svc -A

Show events:

kubectl get events -A

Describe a pod:

kubectl describe pod -n <namespace> podname

As you can see kubectl follows a certain pattern. The usual format is

kubectl (get/describe/delete/edit) <resource-type> -n <namespace>

An exception to this pattern is the logs command:

kubectl logs -n <namespace> <pod-name>

Talosctl

Show talos dashboard:

talosctl dashboard

Show talos kernel logs:

talosctl dmesg