k3s: Document tuning kubernetes image cache

When we aren't using disk on the host for anything else, kubernetes
can simply use as much of the disk as it wants for caching images.
The default for this is upto 85%. However, since we *do* use the disk
for our dind, we want to tune this to take up only half that size.

I've already applied this to the 2i2c-hetzner node

Follow-up to jupyterhub#3229

Author: yuvipanda

docs/source/deployment/k3s.md

@@ -27,13 +27,16 @@ config of _disabling traefik_ that comes built in. We deploy nginx as part of ou
 do not need traefik.
 
 1. Create a Kubelet Config file in `/etc/kubelet.yaml` so we can
-   tweak various kubelet options, including maximum number of pods on a single
-   node:
+   tweak various kubelet options, including maximum number of pods on a single node and when to cleanup unused images:
 
    ```yaml
    apiVersion: kubelet.config.k8s.io/v1beta1
    kind: KubeletConfiguration
    maxPods: 300
+   # Clean up images pulled by kubernetes anytime we are over
+   # 40% disk usage until we hit 20%
+   imageGCHighThresholdPercent: 40
+   imageGCLowThresholdPercent: 20
    ```
 
    We will need to develop better intuition for how many pods per node, but given we offer about