Skip to content

Troubleshooting

Pods not starting after Talos update

One reason can be that images used for kube-system do not match the Talos release. Run talosctl upgrade-k8s to resolve this.

Pod not starting because Jiva volume is already mounted at more than one place

This is rather brutal solution, but it works. I will improve it if I find a better way.

  1. Get <NODE_IP> of node running JivaVolume for PVC: get-pvc-node and then enter PVC name
  2. Reboot the node: talosctl reboot -n <NODE_IP>

Flux can't reconcile a helmrelease

  • Get status of all helmreleases
    flux get helmreleases --all-namespaces
    
    or shorter
    flux get hr -A
    
  • Make sure that sources can be accessed (as there might be typos)
    flux get sources helm -A
    flux get sources git -A
    flux get sources chart -A
    
  • Get more information on helm release (start here when "install retries exhausted" or "Helm uninstall failed: uninstall: Release not loaded")
    kubectl describe helmrelease <RELEASE_NAME> -n <RELEASE_NAMESPACE>
    
  • Restart when Flux gave up
    flux resume hr <RELEASE_NAME> -n <RELEASE_NAMESPACE>
    

Get CPU temperature in Proxmox for Asus PN50

Look at k10temp-pci-00c3 in the output of sensors command.