I run a Nomad cluster of 3 Raspberry pi 4 (2x8Gb, 1x4Gb). All of them booting from SSDs using usb3 to Sata cables. I’m very satisfied with it, I have over 30 workloads running all the time, from simple things like Ansible cron jobs (running in containers, for backing up things) to more resource intensive apps like Plex.
Everything is backed up daily and the entire cluster its configured with Ansible thinking that I was going to have disk/os/hardware failures very soon, but it’s been more that a year running 24/7 with no issues other that a sporadic crash due to OOM that was fixed by adding some swap, can’t complain otherwise.