Delete old images from Talos Linux Kubernetes worker nodes with machine config patch

I ran into an issue where looking over some health metrics on my Talos Kubernetes nodes where I saw disk pressure pretty high on a couple of my worker nodes. It turns out this is fairly easy to get resolved. You can create the following config patch in Omni to apply to your Talos nodes:

machine:
  kubelet:
    extraConfig:
      # Start GC sooner than the default ~85%
      imageGCHighThresholdPercent: 75
      # Keep reclaiming until under this
      imageGCLowThresholdPercent: 65
      # Don't delete brand-new unused images
      imageMinimumGCAge: "2m"
      # Also drop unused images older than 7 days even if disk isn't critical
      # (this is what should eat old gitlab-ee / pulse tags)
      imageMaximumGCAge: "168h"

Apply the config patch on all nodes that have disk pressure and Talos will begin discarding old images immediately. Comparing from before I ran the patch to after: