the one that got isolated looks like it's faling to restore etcd for the embeded rke2 cluster. that caused a bunch of volumes to fault. when i tried to salvage them about 10 of them wouldnt salvage. after messing with it for a while, i ultimately deleted the replica on the failed node, and then i was able to salvage it.

at this point i can either a) remove the node, and the reprovision it, or b) fix the failed node. if you want to make a run at fixing it for documentation purposes, lmk, otherwise i'll nuke it and reinstall it later today.