>@ahrens Currently, zpool_do_labelclear() consults zpool_in_use() which reads
>the labels. If that fails, only with a -f will labelclear proceed anyway. Is
>that good enough, or do we still want -f to have successfully read the labels?
I'd prefer to require some sort of zpool label even with -f. AFAIK the use
case is to make sure that "zpool import" (and related commands, e.g. zdb -e)
don't try to use this device. If there is no zpool label then they already
won't, so there's no need to zero (and no risk of zeroing something that they
didn't intend).
---
Reply to this email directly or view it on GitHub:
https://github.com/openzfs/openzfs/pull/32#issuecomment-152752284
_______________________________________________
developer mailing list
[email protected]
http://lists.open-zfs.org/mailman/listinfo/developer
