Skip site navigation (1)Skip section navigation (2)
Date:      Wed, 15 Apr 2015 12:08:38 +0200
From:      Fabian Keil <freebsd-listen@fabiankeil.de>
To:        freebsd-questions@freebsd.org
Subject:   Re: mountroot prompt with error2, when trying to boot from a single drive in a 2-way mirror
Message-ID:  <05238e5a.6156fafe@fabiankeil.de>
In-Reply-To: <CACo--muDpFszfQZzGe7%2BvuBc1Y9QQ0u-fgoHCM1kOUny2u5yOg@mail.gmail.com>
References:  <CACo--mtMq=pLA-n5B2yPGftdUVa0aR6Pox5mRitMLguPXRto0g@mail.gmail.com> <552D135F.9010907@FreeBSD.org> <CACo--muDpFszfQZzGe7%2BvuBc1Y9QQ0u-fgoHCM1kOUny2u5yOg@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
--Sig_/yjJeiezBcmwPMPL.Yg6/fkk
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: quoted-printable

yudi v <yudi.tux@gmail.com> wrote:

> It's not the BIOS settings, I checked. It picks up the other HDD in the
> mirror and goes through the boot code and then it fails at booting into z=
fs
> root pool.
> The error is:
>=20
> *Trying to mount root from zfs:osysPool/ROOT/default []...
> *>
>=20
> * Mounting from zfs:osysPool/ROOT/default failed with error 6.*
>=20
>  it is something to do with the guid mismatch for ada2p3 and ada3p3, not
> sure why it's even trying to compare them as they are the two partitions =
in
> the mirror.
> Please see the below images for the relevant console messages.
> screen1:
> https://drive.google.com/file/d/1Q-F-8kF-Nevn5ijvFXLNuvtJOuRn7ztO2Q/view?=
usp=3Dsharing
> screen2:
> https://drive.google.com/file/d/1ZGseshS0Uk0cc6Gli_-tywHNXO7sLQ_aVw/view?=
usp=3Dsharing

Please note that these resources aren't accessible without allowing
presumably non-free JavaScript from untrustworthy (YMMV) sources.

> Is anything amiss in the above label info for these two drives?
> I have used these two drives before for testing and reinstalled the os and
> recreated the pools.
> Any suggestions on how to fix this.

The problem could be the result of a known race condition that will prevent
the system from booting if the kernel looks for the root pool before its
vdevs are available. The less disks there are, the "better" the chances that
ZFS "wins" the race.

The vfs.mountroot.timeout is ignored for ZFS so the kernel only tries once.
For details and a patch to change this see:
https://lists.freebsd.org/pipermail/freebsd-fs/2015-March/020997.html

As a workaround you can add a UFS root file system on a disk that
doesn't actually exist to vfs.root.mountfrom. It should be the first
one so you don't hit the spa_namespace_lock deadlock reported in:
https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=3D198563

Fabian

--Sig_/yjJeiezBcmwPMPL.Yg6/fkk
Content-Type: application/pgp-signature
Content-Description: OpenPGP digital signature

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2

iEYEARECAAYFAlUuOKMACgkQBYqIVf93VJ2uSgCgy+7Wl/oEmmJi8ov2bkUQHx+u
O5MAoMAuXxwP/GAuh3SHeDFwIM+PYiXJ
=YlBL
-----END PGP SIGNATURE-----

--Sig_/yjJeiezBcmwPMPL.Yg6/fkk--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?05238e5a.6156fafe>