Skip site navigation (1)Skip section navigation (2)
Date:      Sat, 29 Apr 2006 09:15:02 +0200
From:      Pawel Jakub Dawidek <pjd@FreeBSD.org>
To:        Rong-En Fan <grafan@gmail.com>
Cc:        geom@freebsd.org
Subject:   Re: v_vfs_done error = 1 on gconcat device
Message-ID:  <20060429071501.GA4896@garage.freebsd.pl>
In-Reply-To: <6eb82e0604281734v37c6eca0q8daa5dbca66f58d7@mail.gmail.com>
References:  <6eb82e0604281734v37c6eca0q8daa5dbca66f58d7@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help

--8t9RHnE3ZwKMSgU+
Content-Type: text/plain; charset=iso-8859-2
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Fri, Apr 28, 2006 at 08:34:03PM -0400, Rong-En Fan wrote:
+> Hi Pawel,
+>=20
+> I'm running RELENG_6 around Apr 6. Sometimes, I saw lots of
+>=20
+> g_vfs_done():concat/tmp[WRITE(offset=3D25658605568, length=3D114688)]err=
or =3D 1
+> g_vfs_done():concat/tmp[WRITE(offset=3D25658605568, length=3D114688)]err=
or =3D 1
+> g_vfs_done():concat/tmp[WRITE(offset=3D25658605568, length=3D114688)]err=
or =3D 1
+> g_vfs_done():concat/tmp[WRITE(offset=3D25658605568, length=3D114688)]err=
or =3D 1
+> g_vfs_done():concat/tmp[WRITE(offset=3D25658605568, length=3D114688)]err=
or =3D 1
+>=20
+> in my console. The concat/tmp is consist of two disk:
+>=20
+> Geom name: tmp
+> State: UP
+> Status: Total=3D2, Online=3D2
+> Type: AUTOMATIC
+> ID: 2047271624
+> Providers:
+> 1. Name: concat/tmp
+>   Mediasize: 62055485440 (58G)
+>   Sectorsize: 512
+>   Mode: r1w1e1
+> Consumers:
+> 1. Name: da0s1d
+>   Mediasize: 25658654720 (24G)
+>   Sectorsize: 512
+>   Mode: r1w1e2
+>   Start: 0
+>   End: 25658654208
+> 2. Name: da1s1e
+>   Mediasize: 36396831744 (34G)
+>   Sectorsize: 512
+>   Mode: r1w1e2
+>   Start: 25658654208
+>   End: 62055485440
+>=20
+> If the offset in the console is in the same unit as "Mediasize" in gconc=
at list,
+> then, it looks like it tries to write something that crossing two
+> physical disks.
+> IIRC, everytime I see those messages, the offset is really close to the
+> end of the first disk.
+>=20
+> Any ideas?

It works here:

	# mdconfig -a -t malloc -s 25658654720b -u 0
	# mdconfig -a -t malloc -s 36396831744b -u 1
	# gconcat label foo md0 md1
	# sysctl kern.geom.concat.debug=3D2
	# ./dwrite /dev/concat/foo 25658605568 114688

The console output:

	GEOM_CONCAT[2]: Request received. concat/foo[WRITE(offset=3D25658605568, l=
ength=3D114688)]
	GEOM_CONCAT[2]: Sending request. md0[WRITE(offset=3D25658605568, length=3D=
48640)]
	GEOM_CONCAT[2]: Sending request. md1[WRITE(offset=3D0, length=3D66048)]

I'd more expect that your partitions are not properly configured. Can
you show me the output of "bsdlabel /dev/da1s1"?

PS. # gconcat list
Geom name: foo
State: UP
Status: Total=3D2, Online=3D2
Type: AUTOMATIC
ID: 2742083475
Providers:
1. Name: concat/foo
   Mediasize: 62055485440 (58G)
   Sectorsize: 512
   Mode: r0w0e0
Consumers:
1. Name: md0
   Mediasize: 25658654720 (24G)
   Sectorsize: 512
   Mode: r0w0e0
   Start: 0
   End: 25658654208
2. Name: md1
   Mediasize: 36396831744 (34G)
   Sectorsize: 512
   Mode: r0w0e0
   Start: 25658654208
   End: 62055485440

--=20
Pawel Jakub Dawidek                       http://www.wheel.pl
pjd@FreeBSD.org                           http://www.FreeBSD.org
FreeBSD committer                         Am I Evil? Yes, I Am!

--8t9RHnE3ZwKMSgU+
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.2 (FreeBSD)

iD8DBQFEUxJ1ForvXbEpPzQRAgiPAKCJM27WDNOczHNTgrwrBuA3A/1DDgCg2kuf
Qpp3chSAi4zDpVSDwUDHfB8=
=gHdh
-----END PGP SIGNATURE-----

--8t9RHnE3ZwKMSgU+--



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20060429071501.GA4896>