I have mistakenly copied 10M+ files onto a spinning mirror (copy process took two days, deletion process from old source still going on and probably going to take four days) that is the root dataset of a 2-vdev mirror pool. The mistake is that I did not create a specific dataset in the destination pool before copying. While I know that I can create new datasets in that pool to accommodate for different types of files, is there a way for me to render the current root dataset into a separate child of the root, without re-copying all files and waiting another week? Alternatively, can I destroy the root dataset without affecting its children?
