> On Apr 5, 2016, at 5:56 PM, Jorgen Lundman <lund...@lundman.net> wrote: > >> >> I agree, and "zpool labelclear" is going to clear the label. >> >> "zpool labelclear" is not going to overwrite something that isn't a zpool >> label, in part because I haven't heard any use case for it. > > Wonderful, and the situation where there is just enough of a ZFS label that > "zpool add" and "zpool attach" refuses to work, because there is a pool > there; but not enough of a ZFS label for labelclear to work? This situation > is frustrating.
zpool create, add, and attach have '-f' options to override, at least on Solaris/illumos. There is no need for zpool labelclear in this case. IMHO, labelclear is almost entirely a developer/tester use case, where quickly getting back to uninitialized state is desirable prior to running more tests. > > Now the labelclear conversation is somewhat nit-picking, what is far more > important to us is https://www.illumos.org/issues/6477 > <https://www.illumos.org/issues/6477> > This problem has already blown away users' existing ZFS pools, AND a user's > unrelated HFS filesystem. It worries me :) At the moment we are considering > removing "cache" support on O3X until we can find a solution. agree Although we should not train people to always use -f, which is effectively what people are advocating. There is a damn good reason not using -f should fail. -- richard > > Lund > > -- > Jorgen Lundman | <lund...@lundman.net> > Unix Administrator | +81 (0)90-5578-8500 (work) > Shibuya-ku, Tokyo | +81 (0)80-2090-5800 (cell) > Japan | +81 (0)3 -3375-1767 (home) > ------------------------------------------- openzfs-developer Archives: https://www.listbox.com/member/archive/274414/=now RSS Feed: https://www.listbox.com/member/archive/rss/274414/28015062-cce53afa Modify Your Subscription: https://www.listbox.com/member/?member_id=28015062&id_secret=28015062-f966d51c Powered by Listbox: http://www.listbox.com
