Re: changes to the zfs boot (was: Re: git: 72a1cb05cd23 - main - rc(8): Add a zpoolupgrade rc.d script)
- Reply: Patrick M. Hausen: "Re: changes to the zfs boot (was: Re: git: 72a1cb05cd23 - main - rc(8): Add a zpoolupgrade rc.d script)"
- In reply to: Patrick M. Hausen: "Re: changes to the zfs boot (was: Re: git: 72a1cb05cd23 - main - rc(8): Add a zpoolupgrade rc.d script)"
- Go to: [ bottom of page ] [ top of archives ] [ this month ]
Date: Wed, 09 Nov 2022 20:51:29 UTC
On Wed, Nov 9, 2022 at 12:02 PM Patrick M. Hausen <pmh@hausen.com> wrote: > Hi all, > > > Am 09.11.2022 um 16:54 schrieb Warner Losh <imp@bsdimp.com>: > > >> There is a fixed list of features we support in the boot loader: > > >> [...] > > >> Any feature not on this list will cause the boot loader to > > >> reject the pool. > > I admit that I do not grasp the full implications of this thread and the > proposed > and debated changes. Does that imply that a simple "zpool upgrade" of the > boot/root pool might lead to an unbootable system in the future - even if > the > boot loader is upgraded as it should, too? > Yes. For safety, boot loader upgrade is mandatory when you do a zpool upgrade of the root filesystem. It was definitely needed in the OpenZFS jump, and we've had one or two other flag days since. It would be nice if we had a failsafe here, but we don't today. With a failsafe, we could say 'well, go ahead and try, even if it encounters something it doesn't understand... to at least allow the system to boot. Warner