add nfs-stale-handle checker cronjob

[otosky]

when the Unraid "mover" moves files off cache disk to the disk array, all the inodes change and pods with an NFS mount start to encounter stale file handle issues.

I added a cronjob that checked for these things and would run a kubectl delete pod on affected deployments and statefulsets, but I guess I never put it in the repo.

Add this cronjob version control so that it's covered by flux

[otosky]

this already exists
https://github.com/otosky/home-ops/blob/main/kubernetes/apps/media/restart-pods-with-stale-nfs.yaml