I've now got UPS configured to handle emergency shutdown before power is completely lost and this looks good so far, however assuming that upgrades to the OS will result in a reset of the boot partition and that the UPS client software needs re-installing, is there a hook we can use within Harvester to trigger a patch post update? I'm guessing as the management Rancher cluster is installed on-top of bare metal that this issue wont be a problem for our management cluster as thats not wiped on update as we can manage the OS separately in our case. As mentioned in the post in #C3ASABBD1 we were hoping to find a "correct" way to do this rather than having to directly install - for anyone responsible running a cluster I'd have thought considering power loss to be a fundamental thing in any setup - even in the most power resilient networks things can go wrong...

I've been waiting on this: https://github.com/harvester/harvester/issues/4556

Sounds ideal... will comment

If you know what changes the UPS installation process does you can probably add something to /oem/90_custom.yaml along the lines of https://docs.harvesterhci.io/v1.4/install/update-harvester-configuration