reset should fail more graceful when k0s not installed

[an-tex]

when running k0sctl reset for a host where k0s has not yet been installed leads to the following output:

INFO[0008] ==> Running phase: Gather k0s facts
INFO[0008] ==> Running phase: Reset controllers
WARN[0008] [ssh] xxxx:22: k0s reported failure: command failed: client exec: ssh session wait: Process exited with status 127
FATA[0008] PANIC: runtime error: invalid memory address or nil pointer dereference

when running with debug flag -d it becomes more apparent:

INFO[0008] ==> Running phase: Gather k0s facts
DEBU[0008] [ssh] xxxxx:22: executing `/usr/local/bin/k0s version`
DEBU[0008] [ssh] xxxxx:22: (stderr) bash: line 1: /usr/local/bin/k0s: No such file or directory
DEBU[0008] [ssh] xxxxx:22: executing `test -e /var/lib/k0s/pki/admin.conf 2> /dev/null`
DEBU[0008] [ssh] xxxxx:22: executing `env "KUBECONFIG=/var/lib/k0s/kubelet.conf" /usr/local/bin/k0s kubectl get -n kube-system namespace kube-system -o template={{.metadata.uid}}`
DEBU[0008] [ssh] xxxxx:22: (stderr) env: ‘/usr/local/bin/k0s’: No such file or directory
DEBU[0008] Preparing phase 'Run Before Reset Hooks'
DEBU[0008] Preparing phase 'Reset workers'
DEBU[0008] 0 workers in total
DEBU[0008] ResetWorkers phase prepared, 0 workers will be reset
DEBU[0008] Preparing phase 'Reset controllers'
DEBU[0008] 3 controllers in total
DEBU[0008] ResetControllers phase prepared, 3 controllers will be reset
INFO[0008] ==> Running phase: Reset controllers
DEBU[0008] [ssh] xxxxx:22: draining node
DEBU[0008] [ssh] xxxxx:22: draining node completed
DEBU[0008] [ssh] xxxxx:22: deleting node...
DEBU[0008] [ssh] xxxxx:22: deleting node
DEBU[0008] [ssh] xxxxx:22: executing `stat /run/systemd/system`
DEBU[0008] [ssh] xxxxx:22: File: /run/systemd/system
DEBU[0008] [ssh] xxxxx:22: Size: 80        	Blocks: 0          IO Block: 4096   directory
DEBU[0008] [ssh] xxxxx:22: Device: 19h/25d	Inode: 368         Links: 3
DEBU[0008] [ssh] xxxxx:22: Access: (0755/drwxr-xr-x)  Uid: (    0/    root)   Gid: (    0/    root)
DEBU[0008] [ssh] xxxxx:22: Access: 2023-02-21 10:28:48.558043986 +0100
DEBU[0008] [ssh] xxxxx:22: Modify: 2023-02-21 10:28:48.550043986 +0100
DEBU[0008] [ssh] xxxxx:22: Change: 2023-02-21 10:28:48.550043986 +0100
DEBU[0008] [ssh] xxxxx:22: Birth: -
DEBU[0008] [ssh] xxxxx:22: executing `systemctl status k0scontroller 2> /dev/null | grep -q "(running)"`
DEBU[0008] [ssh] xxxxx:22: leaving etcd...
DEBU[0008] [ssh] xxxxx:22: leaving etcd completed
DEBU[0008] [ssh] xxxxx:22: resetting k0s...
DEBU[0008] [ssh] xxxxx:22: executing `/usr/local/bin/k0s reset`
DEBU[0008] [ssh] xxxxx:22: (stderr) bash: line 1: /usr/local/bin/k0s: No such file or directory
WARN[0008] [ssh] xxxxx:22: k0s reported failure: command failed: client exec: ssh session wait: Process exited with status 127
DEBU[0008] failed to submit telemetry: the client was already closed
FATA[0008] PANIC: runtime error: invalid memory address or nil pointer dereference

[kke]

Yes, panic especially is no good.