Hi Nicolas,
Nicolas Linkert wrote:
I have one thing happening now at boot which must have happened
> during the migration to zfs boot. I get an error message about
> /dev/random: "No randomness provider enabled for /dev/random.
> Use cryptoadm to provide one." Does anyone know how to fix this?
Yep, Constantine spotted this before as well - see Darren's response
at http://www.opensolaris.org/jive/thread.jspa?threadID=28795&tstart=195
Oh, wrt. the sharetab part of that mail, I think the zfs-boot install
script I had wasn't creating a mountpoint for the new sharetab mount
that went in recently, I need to update that at some stage.
Another thing: Is it possible to upgrade to a higher build when
using zfs boot? Is this what LiveUpgrade does? And is there
a step by step instruction on how to use LiveUpgrade with zfs boot?
Unfortunately not yet - I've been simply doing fresh installs
of Solaris, preserving the slices with ZFS pools I want to keep data
from, and then re-setting up my ZFS root install after that.
If you're using the zfs-actual-root-install.sh, and want to install to
an existing pool, you can create your filesystem first, the set the
$ROOT_FS environment variable to the filesystem you'd like to install
ZFS root on (assuming that filesystem is contained in a pool with a
supported configuration)
Eg. after a reinstall/upgrade do:
* remove any old zfs root entries from /etc/vfstab
if you did an upgrade
* zpool import tank, if you did an fresh install
# zfs destroy tank/old-root-fs
# zfs create tank/new-root-fs
# ROOT_FS=tank/new-root-fs
# export ROOT_FS
# ./zfs-actual-root-install.sh foo
Alternatively, you could boot back into ufs-root, perform a LiveUpgrade
of that, and again re-setup your root pool. Not ideal but it should
work.
cheers,
tim
_______________________________________________
zfs-discuss mailing list
zfs-discuss@opensolaris.org
http://mail.opensolaris.org/mailman/listinfo/zfs-discuss
