reFS order of operations

[pufferdude]

I'm in the process of getting my existing Veeam backup repository from NTFS to reFS, but I'm a little unclear on the details. The server is running Win 2016. I know I have to whack the existing repository (and underlying volume) and recreate as reFS... but then I'm not sure how to get my backups working against the "new" repository again. Do I just recreate the repository and edit my backup jobs to use it, and re-run fulls for all jobs? Or is there more to it?

[pufferdude]

Well, looks like you can't delete a repository with backups associated to it... so now I'm guessing you're supposed to point all backup jobs at a different repository in order to "free up" the repository I want to whack, recreate it, then point the jobs at the new repository?

[pufferdude]

Talkin' to myself here, but I seem to be way off the correct path for getting from a NTFS repository to a reFS one. Since I couldn't delete the NTFS repository because jobs were configured for it, I thought maybe I needed to edit the jobs and point them at a temp repository, to allow the deletion of the main one... but when I try to point at a different repository, the job says "move your backups files to this repository first." So clearly I'm missing something (cause that's not gonna happen)... do I actually need to DELETE all of my existing backups (and jobs?) in order to kill the existing NTFS repository and recreate as reFS, or am I missing something obvious?

[foggy]

Hi Jim, here's the procedure for moving to ReFS repository.

[pufferdude]

Thanks! Got it all squared away. My new fulls to reFS seem a LOT faster. Not sure why.. maybe the 64k blocks?