> [2018-10-16 05:39:27] <kevans91> grahamperrin: *nod* open PRs if you hit bugs, CC kevans@
> [2018-10-16 05:41:30] <grahamperrin> I'm sort of waiting for (or to get my head back around) encrypted ZFS then might start from scratch, maybe with a replacement notebook, rolling bectl into the mix.
> [2018-10-16 05:57:05] <grahamperrin> … I sped through the man page for bectl, thanks. If I feel adventurous (at a saner time of day): given my GELI-encrypted boot routine, should I even _attempt_ to create a boot environment?
> [2018-10-16 05:59:21] <grahamperrin> Like, I have a very vague long-ago memory of (beadm) boot environments and GELI being not a good mix. Or something like that.
My first attempt to create a boot environment failed as shown at <https://pastebin.com/R0kw0jnp>. From the head of the paste:
# date ; uname -v
Fri 26 Oct 2018 05:17:39 BST
FreeBSD 13.0-CURRENT r339737 GENERIC-NODEBUG
# bectl create -r 13.0-CURRENT
/ and /boot not on same device, quitting
Documentation: man bectl
The man page should cover, or at least mention, cases such as this.
A Google search for the quit response led me to <https://lists.freebsd.org/pipermail/freebsd-current/2018-August/070984.html> and:
> … for GELI support, FreeBSD 12.0 will not require two separate pools anymore, and there will be migration instructions shortly.
> Allan Jude
Please, are instructions available?
Side note: apologies for the non-wrapped quotes in the opening post. Previews fail <https://redd.it/9rsvoa> and I forgot that it's necessary to manually wrap.
Take. Sometime after this post, I removed the requirement for /boot and / to be on the same device. The following should happen as a result of this post:
1.) Perhaps finally document `bectl -r` in bectl(8), because that's a good solution for posts like this that want to operate on a boot environment for a system that wasn't booted into an existing ZFS BE
2.) Remove the stale /boot and / reference fom libbe(3)